In the transformation of today’s power grids to smart grids, a major challenge to be overcome is to offset the supply-side uncertainties induced by large-scale integration of renewable energy sources. Sophisticated network control and optimization algorithms, in tandem with advances in the technology of energy storage will prove to be essential in successful integration of renewables at large scale. The role of information, feedback, and markets in containing the effects of these uncertainties is, however, perceived to be as significant.
We are working on developing dynamic pricing algorithms and incentive-based market mechanisms, to be implemented by the system operators, to match supply and demand. In one scenario, the system operator collects real-time information about supply and demand and uses this information along with learning and inference algorithms to update a stochastic optimization problem, which minimizes the probability of a large mismatch between supply and demand within a rolling time horizon of finite length. The decision parameters of this optimization problem are real-time prices and/or other quantifiable incentives, which will be then communicated to the consumers. Developing a mathematical model of the consumer behavior and response to price changes is an essential element of this approach and constitutes another aspect of this research. In addition, on the consumer side, we are working on developing algorithms/devices that would autonomously manage consumer’s electricity usage by responding to real-time price signals within the consumers’ preferences while minimizing their average costs.