Artificial neural network application to alleviate voltage instability problem

Abstract

This paper argues that the artificial neural network (ANN) is an important technique and can be applied to control voltage instability problems in two ways. The first is to control voltage collapse by using load shedding. In this way the minimum and optimal ratio of load shedding can be calculated. The second is to calculate the reactive power required to control sources in the electric power system. The strengths of this powerful technique lie in its ability for modeling and solving many types of problems. The ANN is designed for those two mentioned ways. A multi-layer feed forward ANN trained with error backpropagation learning is proposed. This network is applied to a stressed power system at different load levels. Simulation results on a test system are reported in this paper.