BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Plug-and-Play Operation of Microgrids - Florian Dorfler\, ETH Zuri
 ch
DTSTART:20141127T140000Z
DTEND:20141127T150000Z
UID:TALK56372@talks.cam.ac.uk
CONTACT:Tim Hughes
DESCRIPTION:Microgrids are low-voltage electrical networks composed of dis
 tributed generation\, storage\, load\, and managed autonomously from the l
 arger utility grid. Modeled after the hierarchical control architecture in
  bulk power systems\, a layering of primary\, secondary\, and tertiary con
 trol has become the standard operation paradigm for microgrids. Despite th
 is superficial similarity\, the control objectives in microgrids across th
 ese three layers are varied and ambitious\, and they must be achieved in r
 eal-time without time-scale separations\, without centralized decision mak
 ers\, and ideally in a model-free fashion. In this seminar\, we explore di
 fferent control strategies for these three layers and illuminate some poss
 ibly-unexpected connections and dependencies among them. We build upon dec
 entralized primary droop control strategies and motivate the need for addi
 tional secondary regulation. We find that distributed averaging-based seco
 ndary control architectures using communication among the generation units
  offer the best combination of flexibility and performance. We further lev
 erage these results in tertiary-level energy management tasks such as an e
 conomic generation dispatch. Surprisingly\, we show that the minimizers of
  the economic dispatch optimization problem are in one-to-one corresponden
 ce with the set of steady-states reachable by droop control. This equivale
 nce results in simple guidelines to select the droop coefficients\, which 
 include the well known criteria for power sharing. Finally\, we stress the
  idealistic assumptions underlying any droop-based control architecture an
 d propose a virtual oscillator control paradigm that contains droop contro
 l in the asymptotic limit. We illustrate the performance and robustness of
  all of our designs through hardware experiments.
LOCATION:Cambridge University Engineering Department\, LR5
END:VEVENT
END:VCALENDAR
