Main Article Content

Authors

In this work, a black-box model for the estimation of the rotor angle of a synchronous generator is developed by using voltage and current measurements at the terminals. The model is based on neural networks of the multilayer perceptron (MLP) type. From time-domain simulations for a basic power system, which consists of a generator connected to an infinite bus bar, voltage and current records are collected, which are used as a database to train and validate the neural network that is proposed for the estimation of the rotor angle. It is found that the neural-network based model adapts itself very well to the classic model of the generator, exhibiting a mean square error of 1´10-10 . The results show the validity of the estimation method and foster its potential use for stability studies of the synchronous generator.

René A. Barrera

escuela de ingenieria electrica,electronica y de telecomunicaciones,universidad industrial de santander,bucaramanga,colombia

Juan J. Mora

programa de ingenieria electrica,universidad tecnologia de pereira,pereira,colombia

Sandra M. Pérez

programa de ingenieria electrica,universidad tecnologia de pereira,pereira,colombia
1.
Barrera RA, Mora JJ, Pérez SM. neural estimator of the rotor angle of synchronous generators from measurement of voltage and current at terminals. inycomp [Internet]. 2009 Jan. 9 [cited 2024 Nov. 22];11(1):9-20. Available from: https://revistaingenieria.univalle.edu.co/index.php/ingenieria_y_competitividad/article/view/2467