Accurate pressure prediction of a servo-valve controlled hydraulic system

KILIÇ E., DÖLEN M., KOKU A. B., ÇALIŞKAN H., Balkan T.

MECHATRONICS, vol.22, no.7, pp.997-1014, 2012 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 22 Issue: 7
  • Publication Date: 2012
  • Doi Number: 10.1016/j.mechatronics.2012.08.001
  • Journal Name: MECHATRONICS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.997-1014
  • Keywords: Hydraulic system, Pressure dynamics, Nonlinear system modeling, Long-term prediction, Structured neural network, NEURAL-NETWORK, IDENTIFICATION, MODELS
  • Middle East Technical University Affiliated: Yes

Abstract

The main goal of this paper is to predict the chamber pressures in hydraulic cylinder of a servo-valve controlled hydraulic system accurately using advanced modeling tools like artificial neural networks. After showing that the black-box modeling approaches are not sufficient for long-term prediction of pressures, a structured neural network model is proposed to capture the pressure dynamics of this inherently non-linear system. The paper shows that the proposed network model could be easily trained to predict the pressure dynamics of an experimental hydraulic test setup provided that the training session is initiated with the weights of the developed model. (C) 2012 Elsevier Ltd. All rights reserved.