Voltage ranking using artificial neural network

Voltage ranking attempts to rank busbar voltage deviations from their norma lly accepted security margins based on a set of performance indices (PI), w ithout performing a full load flow. Existing methods suffer from either mas king effects or long computation time. In this paper, an artificial neural network method is proposed for voltage ranking Counterpropagation network ( CPN) has been employed to overcome the problems listed above. A variety of input features are used with the aim of lowering the dimension of the propo sed ANN to make it applicable for large power systems. The method is tested on two example systems, a five-bus system and a 71-bus system with very en couraging results.