A Comparative Study on Static Voltage Stability Analysis of Power System using Feed Forward Neural Network and Radial Basis Function Neural Network with Reduced Set of Inputs

Power systems are being operated close to their voltage security limit which makes their reliable operation more challenging. Voltage stability has been a major problem faced by utilities now a days. Many blackout have been reported around the world related to voltage instability problem. Many voltage stability analysis methods have been presented up till now. Here artificial neural network based approach for voltage stability analysis of power system is presented. Voltage stability analysis of power system has been done in terms of voltage stability margin. Voltage stability margin is calculated by estimating distance from current operating point to maximum voltage stability point according to loading parameter. Principal Component Analysis is an efficient feature selection method used here to find few input variables required to estimate voltage stability margin with sufficient accuracy and execution speed. Major drawback of previously published work is they need to train new neural network whenever change in system topology occur which is very time consuming process. Therefor in this paper single ANN has been prepared for estimating VSM for several system configuration. Feed forward neural network and radial basis function neural network have been used for voltage stability analysis of power system. The final result indicate that proposed method provide an efficient and compact ANN model for different system configuration as well as base case operating condition with lowest possible input required.