#INFORMS15 TutORial 3 (today 1:30pm): Uncertainty in demand response – identification, estimation and learning

This tutorial is by Joshua Taylor of the University of Toronto and Johanna Mathieu of the University of Michigan in Room 108A today Sunday at 1:30pm.


Demand response from flexible electric loads such as electric vehicles, air conditioners, and smart home appliances represents a vast, clean, and potentially high-performance resource for the electric power system. Compared to traditional resources like generators, transmission lines, and storage devices, loads are highly uncertain. While effective techniques have been developed over the past century for accommodating load uncertainty in power system operations, new techniques are required to enable loads to become a reliable and effective resource for the power system operator. Much of this uncertainty falls under the purview of the load aggregator, whose job it is to provide power system services like load shifting, curtailment, and regulation by managing the power consumption of individual members of a load population. A load aggregator may not know the state or model of a load and furthermore may not be able to directly measure the loads’ responses to its commands.

In this tutorial, we survey techniques for managing load uncertainty in demand response. Our discussion is organized into three problem types: identifying load models, estimating load states, and learning these features in conjunction with deploying the loads for demand response. We also discuss new challenges and future directions for each problem type.


Joshua A. Taylor received the BS degree in 2006 from Carnegie Mellon University and the SM and PhD degrees in 2008 and 2011 from the Massachusetts Institute of Technology, all in Mechanical Engineering. From 2011 to 2012, he was a postdoctoral researcher in Electrical Engineering & Computer Sciences at the University of California, Berkeley. He is currently an Assistant Professor of Electrical and Computer Engineering at the University of Toronto. His research interests include renewable energy and demand response, machine learning, and infrastructural couplings.

Johanna L. Mathieu received the B.S. degree in ocean engineering from MIT in 2004 and the M.S. and Ph.D. degrees in mechanical engineering from the University of California, Berkeley in 2008 and 2012, respectively. She is currently an Assistant Professor in the Department of Electrical Engineering and Computer Science at the University of Michigan. Prior to joining the University of Michigan, she was a postdoctoral researcher at ETH Zurich, Switzerland. Her research focuses on methods to model and control distributed demand response and energy storage resources to improve power system reliability and support the integration of renewable energy resources.