Voltage stability assessment in power systems using Artificial Neural Networks

Voltage stability is a major concern in planning and operations of power systems. The main factor causing instability is the inability of the power system to meet the demand for reactive power. In order to operate power system with maximum security and reliability, knowledge of the voltage stability margin is of vital importance to utilities. L-index is used as the indicator to voltage instability. Performance of an Artificial Neural Network (ANN) is done using MATLAB Neural Network toolbox. Using PSAT (Power System Analysis toolbox) IEEE-14 bus system power flow analysis is done. L-index is calculated. With 4 inputs and L-index as output ANN is trained and performance is analysed. Error between the actual and predicted output was found to be small. It is found that ANN can efficiently predict the voltage for new values of inputs.