Study of different ANN algorithms for weak area identification of power systems

This paper presents the suitability of different artificial neural network (ANN) algorithms in estimating the voltage instability of power systems. The ANN models based on different training algorithm are designed and a comparative study is carried out to accurately predict the voltage collapse phenomenon. In the present study, L-index is used as the voltage collapse proximity indicator. This approach is tested on a sample 5-bus system taken from the literature. It is found that the results obtained are quite promising in predicting the voltage collapse phenomenon.