NEURAL-NETWORK-INVERSE-BASED TCSC CONTROL STRATEGY USING ONLY LOCAL MEASUREMENTS

戴先中,张凯锋,陆翔

DOI: https://doi.org/10.3321/j.issn:0258-8013.2004.07.007

2004-01-01

Abstract:Using only local measurements axed taking measurable active power passing through TCSC as the control objective, this paper presents a novel neural-network-inversebased TCSC control strategy. Based on TCSC's detailed nonlinear dynamic model, TCSC system is strictly linearized, and the proposed method can be applied to arbitrary power system topology. Neural-network-inverse-system method is used to construct the controller, and thus the proposed strategy is not dependent on the system's accurate model axed parameters. The simulation results aiming at a two-area four-machine power system demonstrate that the both goals of directly controlling active power through TCSC and indirectly controlling remote generator's power angle can be simultaneously reached.