Neural-Network-Based Wide-Area Damping Control Considering Communication Delays

Abstract

This paper proposes an improved neuro-adaptive control scheme, based on online system identification and simultaneous control, to replace a residue-based wide-area damping controller for static var compensator for low-frequency oscillation damping. Wide-area signals, required for effective damping of inter-area oscillations, introduce practically unavoidable time-delays that can offset the effectiveness of a controller and therefore delay compensation is required for such applications. The proposed controller comprises of a simple, linear neural identifier, with a few adjustable connection weights which ensures minimal computational burden and fast learning capability. The parameters of the controller, designed using optimal control with one-sample-ahead output prediction, are related to the identifier parameters. Far superior oscillation-damping performance over a wide range of communication delays, in comparison with a well-designed residue-based controller, has been validated through simulation studies on an interconnected power system. Moreover, the proposed control scheme does not require any compensation for the time-delay.