Static Security Assessment of a Power System Using Genetic-Based Neural Network

Abstract

Secure operation of modern electric power system is important despite of occurrence of various disturbances in the system. Assessing the security status is a computationally demanding task, and there is a need for a fast on-line security assessment tool. To achieve this, an artificial neural network (ANN) approach to power system static security assessment is proposed with the objective of determining the suitability of using ANN on a large-sized power system. A modular ANN design approach is proposed in which separate ANNs are dedicated to handle the security assessment task with respect to different areas of a power network. The developed ANNs are trained to classify the secure and insecure states of a power system for a set of specified network and power contingencies. The proposed method emphasizes the use of genetic algorithm for determining the optimum values of the ANN parameters, such as the momentum, learning rate, and hidden neurons. The genetic-based ANN approach has been implemented on the Malaysian power system. By implementing the genetic-based ANN, the selection of the ANN parameters by trial and error method can be avoided, and results show that the accuracy of the genetic-based ANN method is better than that of the nongenetic–based ANN.