BDBComp
Parceria:
SBC
A Proposal of an Intelligent Forecasting System for Automotive Diagnostics using Time Series Analysis

Diogo M. AlmeidaDaniel C. CunhaPaulo S. G. de Mattos Neto

Automotive electronics and the use of software for control and monitoringhave been increasing significantly in the design of modern vehicles andtheir on board diagnosis (OBD) systems. A desirable attribute of an OBD systemis the ability to make fault predictions to avoid major disruptions and damagesto the driver. Thus, the objective of this paper is to propose a hybrid forecastingsystem for automotive diagnostics that is more robust than OBD system and isa combination of a linear - autoregressive integrated moving average (ARIMA)model and a nonlinear - artificial neural network (ANN) model in order to overcomethe accuracy and limitations of each model separately. The proposed approachshowed superior results than single models in five evaluation measures.

http://www.lbd.dcc.ufmg.br/colecoes/eniac/2016/049.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD