State estimation and PEV scheduling in smart grids

This dissertation studies state estimation (SE) and scheduling algorithms for modeling and managing smarter electric power systems. SE is the backbone of control functions in the energy management system of the power grid. In this dissertation, we consider SE at both the transmission system and the distribution system. Although SE has been widely implemented in the transmission systems, deployment of a large number of phasor measurement units (PMU) call for major changes in the legacy SE algorithms. This dissertation presents a reduced-order forecasting-aided state estimator which effectively incorporates PMU measurements with conventional measurements and demonstrates its viability through simulations.;The evolution of the grid towards a smarter grid will mean an increasing demand for situational awareness and require SE at the distribution level. We develop an enhanced forecasting-aided state estimation (FASE) algorithm that incorporates forecasting-aided topology change detection and an event-triggered recursive Bayesian estimator to identify the correct topology of the distribution network. Simulation studies with microgrid-induced changes are presented to illustrate the effectiveness of the proposed algorithm.;This dissertation also considers the problem of optimizing the charging of plug-in hybrid electric vehicles (PHEVs) with a centralized controller. As the penetration level of PHEVs increases, their charging will become more stressful to the distribution system. In our studies here, we concentrate on PHEV charging in public infrastructures, such as shared parking lots in commercial office campuses and shopping malls. In such clustered scenarios, PHEV charging demand can vary significantly and stochastically with time, and the overall system performance is a trade-off between the peak load on the distribution system and the consumer specified deadlines that are met successfully. In these scenarios, centralized scheduling of charging will be beneficial, unlike distributed residential charging. In this dissertation, we show that centralized scheduling algorithms can help the parking lot owner to balance distribution system load with quality of charging service.