Smart Grid Seminar
Achieving fairness for EV charging in overload: a fluid approach
When dimensioning the power capacity of an Electrical Vehicle charging garage, it may not be practical to provision for the peak load (all chargers operating at full power). Instead, some scheduling/curtailment policy can be put in place to manage the available capacity, exploiting users' statistical multiplexing and their deferability of service.
In this talk we present a fluid model for the car population, in the form of a PDE in service-time/sojourn- time coordinates. Through this model we analyze the performance of different scheduling policies. We show that in overload, the total amount of unfinished work is the same for all policies, but they differ in the distribution of work across users. With fairness in mind we introduce a new policy called Least Laxity Ratio, which achieves proportionality in reneged work at the fluid scale. We also validate the empirical performance of the various policies by simulation, including experiments with realistic data sets.
BIO: Fernando Paganini received his PhD in EE from Caltech (1996), was a postdoc at MIT (1996-1997), and a Faculty member at UCLA EE (1997-2005). Since 2005 he is Professor of Engineering at Universidad ORT, Uruguay. He has received several awards, including the 1995 O Hugo Schuck, the 1996 Caltech Wilts and Clauser Prizes, the 1999 Packard Fellowship, and the 2004 George S. Axelby award. He is a member of the Latin American Academy of Sciences and a Fellow of the IEEE.
Contact: Daniel Guo email@example.com