Complexity, Approximation, and Relaxation of the Power Flow Equations

Wednesday, January 28, 2015
12:00 PM to 1:00 PM
ART 1.102
Free and open to the public

The design, control, and operation of the power grid, probably the largest and most expansive system ever engineered, require the solving of optimization problems over the steady-state power flow equations. The resulting mixed nonconvex programs are often computationally challenging and increasingly so with the increased stochasticity in generation and load. This talk presents some new complexity results, as well as a number of advances in approximating and relaxing the power flow equations to address emerging applications in power systems, including large-scale power restoration after blackouts, the design of resilient networks, and the integration of renewable generation. Extensive computational results demonstrate some of the benefits of the proposed techniques.

x x


Pascal van Hentenryck

Pascal van Hentenryck

Australian National University

Pascal Van Hentenryck leads the Optimisation Research Group (about 75 people) at National ICT Australia (NICTA). He is the recipient of two honorary degrees and is a fellow of the Association for the Advancement of Artificial Intelligence. He was awarded the 2002 INFORMS ICS Award for research excellence in operations research and computer science, the 2006 ACP Award for research excellence in constraint programming, the 2010-2011 Philip J. Bray Award for Teaching Excellence at Brown University, and was the 2013 IFORS Distinguished speaker. Van Hentenryck is the author of five MIT Press books and has developed a number of innovative optimisation systems that are widely used in academia and industry. Van Hentenryck's research is currently at the intersection of data science and optimization with a focus on disaster management, energy systems, recommender systems, and transportation.