On-line static security assessment of power systems by a progressive learning neural network

In this paper an application of artificial neural networks to the static security assessment is presented. The main feature of this approach is the on-line learning of the relationship between the system operating point and the security related variables. A clustering neural network, purposely developed for on-line learning of a continuous data flow, is employed. Two different methods, aimed respectively to contingency ranking and direct post-fault values prediction, have been devised and compared. Numerical tests, carried out on the IEEE-30 bus system by employing a simulator purposely set up, are presented and discussed.