Position: PhD Student
Current Institution: Massachusetts Institute of Technology
Abstract: Toward Economically-Efficient and Technically-Realizable Electric Energy Service
The proliferation of intermittent, renewable technologies combined with the customer service-oriented trends present new fundamental challenges. Given temporal criticality of operating the physical grid, it is essential to process vast sources of asynchronously gathered data online and use for multi-layered interactive automation at well-understood expected performance. Today’s Supervisory and Data Acquisition (SCADA) architecture capable of enabling electricity service is limited to hierarchical temporal spatial assumptions that no longer hold. SCADA must evolve into flexible architecture capable of aligning end-to-end physical and economic processes. Our approach provides a means of multi-layered modeling of interactions in terms of commonly-understood energy and power variables characterizing multi-physical energy conversion processes. This modeling framework allows for the zooming in and out of aggregate models as per the spatial and temporal granularity of interest. The most novel finding so far: such multi-layered modeling with intra- and inter-layer interactions not only eases the component-level control design, but also results in a linear interactive model for aggregate-level analysis, thus allowing for end-device participation at value. The approach uses formal mathematical techniques called spatial and temporal lifting that facilitate abstraction of inner component details, revealing only the interface variables (energy and power) to its neighbors and its time-varying limits to its coordinator as a function of economic signals that are learned by the component itself or communicated by its coordinating aggregator. Finally, this framework induces a method for systematic practical composition of an otherwise complex computer platform for simulating technical and economic information exchange within a future electric-energy system. Ultimately, the modeling framework-supported control design demonstrating the effects of end-to-end market transactions on the power grid is being shown using our home-grown scalable electric power system simulator (SEPSS) to result in efficient and technically realizable integration of renewables and near-optimal provision of electricity services.
Rupamathi Jaddivada is currently a third-year PhD student at MIT, supervised by Professor Marija Ilic in the Laboratory for Information and Decision Systems (LIDS). She received a bachelor’s degree in electrical and electronics engineering from Jawaharlal Nehru Technological University in India in 2014, and a master’s degree in electrical engineering and computer science from Carnegie Mellon University in 2015. She interned at the New Electricity Transmission Software Solutions in 2018, working on developing software modules for efficiency enhancements in power grid operations. She received a gold medal in the Srinivasa Ramanujan Mathematics competition (SRMC) in 2014. Her research interests include modeling control and numerical simulations of complex dynamical systems, in particular for electric power system applications.