Periodic Steady-State Analysis of Large-Scale Electric Systems Using PoincarÉ Map and Parallel Processing

Abstract

An advanced time-domain representation suitable for accelerated periodic, steady-state solutions of large-scale electric power systems is introduced in this paper. The power system, which is three-phase in nature, is modeled by a set of ordinary differential equations. In addition to the all-important transmission lines' geometric imbalances, frequency-dependency and long-line effects, the transformers' saturation characteristics are also incorporated. In order to provide realistic initial conditions, the data needed for the modeling of the electric system is obtained from power-flow studies. The time domain solution of the overall set of differential equations is computed very reliably using a powerful blend of Newton methods and parallel processing techniques. Whereas the computation of the periodic steady-state solution is obtained with an acceleration procedure based on Newton methods and the Poincare/spl acute/ map, the application of parallel processing techniques using multithread programming speeds up further the time taken by the acceleration process to get to the periodic, steady-state solution. To show the effectiveness and versatility of the newly developed environment, transient and steady-state analysis are carried-out for a three-phase version of the IEEE 118-node system, where nonlinearities are incorporated in the form of the transformers' saturation characteristics and a static var compensator.