Metodología para la sintonización de parámetros en modelos de sistemas de control de centrales eléctricas. 169 hojas. Quito : EPN.

Abstract

To have duly adjusted and calibrated models of the elements of the electric power system, which allow electrical studies of any kind to be carried out on a reliable database, the present article proposes a methodological alternative for the tuning of parameters in the mathematical models of control systems of power generation plants, by using the algorithm of meta-heuristic optimization of particle swarm (PSO - Particle Swarm Optimization). Using COMTRADE (Common Format for Transient Data Exchange) files of an Automatic Disturbance Recorder, the values of Voltages and Currents of events raised in the area of electrical influence of a plant are extracted. These events are simulated as electromagnetic transients (EMT) in the equivalent network structured in DIgSILENT PowerFactory, recreating the conditions in which they happened. Finally, the PSO algorithm developed in DIgSILENT Programming Language (DPL), is implemented and executed in the equivalent network to find the parameters of the control systems of the Esmeraldas II plant, which allow minimize the difference between the results of the simulations and COMTRADE file data.