April 30, 2015
EE Conference Room
Speaker: Konstantin Turitsyn, Assistant Professor, Mechanical Engineering Department, MIT
Dynamic Security Assessment is one of the most challenging computational problems in power systems. Large size of power systems and immense variety of possible contingency scenarios make the brute-force simulation approaches overly prohibitive for practical purposes. This talk will review possible alternatives to direct analysis methods. The first part will focus on the new iterative pruning techinique for fast selection of dangerous N-2 contingencies, and statistical characterization and classification of the observed contingency sets. In the second part, the problem of transient stability analysis problem. The best known approaches to this problem known under the name of direct energy methods in power systems community often suffer from conservativeness and poor scalability. I will present a novel method of stability certification based on the construction of nonlinear Lyapunov functions that decay in some neighborhood of operating point. These Lyapunov function form a convex set defined by linear matrix inequalities which allows adaptation of the certificate to the most common contingencies. Moreover, the construction of the certificate can be accomplished in polynomial time and is tractable even for large scale systems. The key elements of the Lyapunov function construction and and possible applications of the method will be discussed in the end of the talk.
Konstantin Turitsyn received the M.Sc. degree in physics and applied math from Moscow Institute of Physics and Technology in 2004 and Ph.D. degree in physics from Landau Institute for Theoretical Physics, Moscow, in 2007. Currently, he holds a Skolkovo Foundation Career Development Assistant Professor position in the Mechanical Engineering Department at Massachusetts Institute of Technology (MIT), Cambridge. Before joining MIT, he was an Oppenheimer Fellow at Los Alamos National Laboratory working with the Smart Grid research group. His research interests encompass a broad range of problems related to development of novel mathematical tools for analysis of complex nonlinear and stochastic systems. These tools have been applied to problems arising in different domains, most importantly in the fields of statistical physics, optics, as well as mechanical and power engineering.
Hosted by Javad Lavaei.