The objective of this research is to develop a high-performance novel and intelligent grid optimization architecture which is distributed (low-cost) and that balances supply and demand in real-time over multiple levels of the electric power grid. The approach is to link power flow optimization routines across a three-layer hierarchy of power grid: circuits, substations, and the transmission grid using the proposed integrated Approximate Dynamic Programming and Dynamic Stochastic Constraint Optimal Power Flow methodology in order to meet the requirements of dispatching unit portfolios with different objectives at different layers of hierarchy.
Intellectual Merit: The key merit of this project is the ability of the proposed method to accommodate high in-feed of distributed resources and storage, provide a low cost solution to exponential increase in decision and grid variables and allow generation schedule dispatch with low bidding margin, including both supply and demand elasticity. This is possible by implementing massively distributed intelligent optimization architecture with coordinated hierarchical and distributed control that can interact with grid devices. This approach will allow the transmission and distribution network to operate efficiently satisfying various grid level objectives simultaneously and yet achieving global optimality conditions.
Broader Impact: The broader impact is the ability of the proposed infrastructure to transform the power grid functionalities that has substantial economic (improved grid efficiency and optimal supply and demand elasticity), societal (sustainable, reliable and resilient infrastructure) and educational value (new curriculum methods and generation and retention of next generation global workforce) which is important for developing next generation power grid.
