Classification of Faults in Switch Machine using Type-1 and Non-singleton Fuzzy Logic System Trained by Hestenes and Stiefel's Conjugate Gradient Method
Alexandre Queiroz Zamagna Bouhid1; Renan Piazzaroli Finotti Amaral1; Leonardo Goliatt da Fonseca1; Eduardo Pestana de Aguiar1
1 Federal University of Juiz de Fora
doi:10.20906/CPS/CILAMCE2017-0755
Resumo
This paper discusses and analyzes the performance of a technique for classifying possible faults (lack of lubrication, lack of adjustment and malfunction of a component) that can occur in an electromechanical switch machine, which is an equipment used for handling railroad switches. Aiming to better classify these faults, the type-1 and non-singleton fuzzy logic system trained by Hestenes and Stiefel's conjugate gradient method is presented. The performance of the proposal is compared with four classifiers reported in the literature (Bayes based, multilayer perceptron neural network, type-1 and singleton fuzzy logic system trained by steepest descent method and type-1 and singleton FLS trained by the Conjugate Gradient method), demonstrating higher accuracy and faster convergence speed. Based upon the attained results, the proposed model enables the railway company to adopt solutions to achieve operational excellence.
Palavras-chave: Fuzzy Logic System; Non-singleton; Classification; Conjugate Gradient Method; Optimization