Artificial neural net based stability study of power systems
D.R. Marpaka, Michael H. Thursby, S.M. Aghili · 2002
A novel approach for studying the transient stability of a power system comprising a single machine connected to an infinite bus using an artificial neural network (ANN) is presented. This approach uses a multilayer ANN for determining the critical clearing time for the system under consideration. The network was trained with the angular positions and angular velocities up to the last switching operation and the corresponding state of the system as the desired output. After the network obtained the capability of determining the state of the system for different angular positions and velocities, it was presented with a set of data representing the operating conditions, not presented during the training, for determining the critical clearing time for the system under consideration. An example of a power system is presented to illustrate the method.>