Optimization and computation applied to power system scheduling and control

Abstract

Algorithms have been developed for dynamic rescheduling and control of generation to modify the state of an interconnected power system. The algorithms can be used to achieve control of m-1 scalar quantities in a power system by individually adjusting the generation at m power plants. Specifically, the algorithms can be used to schedule the power flow on individual lines or tie-lines to other systems, or the voltage phase-angle differences between arbitrary nodes in a power system. The algorithms developed are for use in a preventive mode, and would be used to improve power system security. The rescheduling problems are solved by linearization and utilize techniques of optimal control theory. The algorithms developed have computational requirements, when advantage is taken of the sparsity of the matrices involved, which make possible their utilization as on-line tools. The control, used to move the generation to the newly rescheduled state, is also quite easy to implement and can be carried out by a central computer facility. The control approach described yields very smooth transitions in power output between the existing and rescheduled levels. Simulation results demonstrate that control of the system to the newly scheduled state can be accomplished in about 2 min. © 1971.