Static security assessment using artificial neural network

Deregulation of power system in recent years has turned static security assessment into a challenging task for which acceptably fast and accurate assessment methodology is essential. Occurrences related to over and undervoltage and line overloading have been responsible for undesirable power system collapse leading to partial or even complete blackouts. This paper presents a research work on artificial neural network (ANN) to examine whether the power system is secured under steady-state operating conditions. The ANN gauges the bus voltages and the line flow conditions. Using the method, detailed load flow study is can be omitted provided that the data supplied to ANN sufficiently cover these operating constraints. A methodology using minimum number of cases from the available large number of contingencies in terms of their impact on the system security has been developed. For training, data from Newton Raphson load flow analysis are used. The artificial neural network has been developed using multilayer feed forward network with backpropagation algorithm. The input variables to the network are loadings of the lines and the voltage magnitude of the load buses. The algorithms are initially tested on the 5 bus and verified on the IEEE-14 bus test system. The results obtained from both test systems indicate that ANN method is comparable in accuracy to the Newton Raphson load flow method with enhanced computational time taken in the process.