Application of ANN methods for instrument transformer correction in transmission line protection

The aim of this paper is to present ANN methods intended for dynamic compensation of current transformers (CTs) and capacitive voltage transformers (CVTs) used for measurements in high voltage power systems. The nonlinear multilayer recursive artificial neural networks (ANNs) have been used for modelling inverse transfer functions of the transformers. The problems as well as the results obtained for different network parameters are presented. The issue of correct application of these novel methods in real protective relays has been discussed.