NEURAL NETWORK BASED SECURITY ASSESMENT OF POWER SYSTEMS

Abstract

The active and reactive components of power flow vector together constitutes the operating point of a power system. If the criteria of security is the prevention of line overloads, the boundaries of the secure domain of the state space are given by the maximum acceptable currents of the transmission lines. In this dissertation work the concept of an Artificial Neural Network using Kohonen's self-organizing feature maps for classifying the states of power system into secure or insecure domain has been used. The idea is based. on the assumption that information in the brain is stored on a two dimensional surface, and that related information occupies neighboring locations on• that surfate. This classifier maps vectors of an N - dimensional space to a two dimensional neural net in a non-linear way preserving the topological order of the input vectors. Hence, secure operating states are the vectors inside the boundaries of the secure domain mapped to a region of neural map different from the region of insecure operating points. A non-linear power system model has been used to study these mappings.