TOWARDS A NEURAL-NETWORK-BASED VOLTAGE STABILITY ASSESSMENT

This paper presents a neural-network based method for evaluating on-li ne power system voltage stability starting with an appropriate lineari zed model of the system, the proximity of the power system to the volt age instability is quantified by using a suitable index. Then, a three -layer feedforward neural network is trained to learn the underlying f eatures of the voltage stability phenomenon. Once trained, the neural network provides the above mentioned voltage stability index as output to a pre-defined set of input variables that are known to influence d irectly the voltage stability conditions of the load area. The effecti veness of the proposed approach has been tested on the 39-busbar New E ngland test system.