Research and Development of a Methodology for the Emergency State Control for Electric-Energy Systems. Final Report.

摘要

This report presents the development of a stochastic approach to solving the emergency state control problem in electric power systems. It identifies slow speed dynamics and cascading failures as characterizing many major electric power system collapses and presents a probabilistic representation of the uncertain changes in power system's configuration. The control algorithm looks several time steps ahead to account for system uncertainty and is aimed at guiding the system from emergency state back into a normal state in an optimal fashion. Algorithms for performing the constrained nonlinear optimization arising from the control computation are reviewed and a state-of-the-art multi-purpose nonlinear optimization code is recommended. The main contribution of this research project is that the uncertainty in power system dynamics is clearly addressed and the emergency state control problem is formulated as a stochastic control problem to reflect the true nature of the system. Furthermore, the stochastic control problem is solved using a sequential decision tree strategy with manageable computational requirements. (ERA citation 09:000658)