Soumydip Ghosh Business Analytics and Math Sciences IBM TJ Watson Research Center Friday, September 28, 2012 An Applied Mathematics Perspective on the Smart Grid The energy infrastructure faces systemic challenges throughout the whole “supply-chain” of generation, transmission and distribution with the widening adoption of new energy technologies such as renewable generation, distributed storage and (vehicle) charging. Smart Grid, broadly defined as new algorithms to control existing infrastructure as well as new infrastructure to facilitate two-way digital communication and signaling between all significant energy resources and consumers down till individual households, offer unique opportunities in meeting these challenges. Mathematical problems that arise in this space include efficient algorithms for day-ahead unit-committment planning and economic dispatch under uncertainty, statistical estimation of energy consumption characteristics under dynamic incentives, dynamic schemes for demand-response based mediation of energy supply and demand, and distributed algorithms for energy dispatch and electrical vehicle charging. The talk will present an overview of some of these problems, focusing on why they are important to stakeholders/end-users as well as are interesting as research questions, and will present a summary of recent results. The talk is derived from our group’s efforts in the area of smarter energy research and practice. Part of this work is being funded by a DoE grant on demonstrating a Smart Grid implementation over a part of the Pacific Northwest grid.