A hybrid modeling using clustering algorithm for textile slashing process

Authors:
Zhang Yuxian;Liu Min;Wang Jianhui;Wang Dan;Ma Yunfei
Affiliations:
Department of Automation, Tsinghua University, Beijing, China;Department of Automation, Tsinghua University, Beijing, China;Faculty of Information Science and Engineering, Northeastern University, Shenyang, China;Faculty of Information Science and Engineering, Northeastern University, Shenyang, China;Faculty of Information Science and Engineering, Northeastern University, Shenyang, China
Venue:
CCDC'09 Proceedings of the 21st annual international conference on Chinese control and decision conference
Year:
2009

Citing 3
Cited 0

Neural networks for control systems: a survey

Automatica (Journal of IFAC)
Original Contribution: Stacked generalization

Neural Networks
An intelligent approach to integration and control of textile processes

Information Sciences: an International Journal - Special issue analytical theory of fuzzy control with applications

Quantified Score

Hi-index	0.00

Visualization

Abstract

The slashing is a very important procedure in textile manufacturing process which can improve warp quality, loom efficiency and reduce warp break. A hybrid modeling method is proposed for textile slashing process. Data are divided to multiple subsets by clustering algorithm, and then artificial neural networks (ANN) and partial least square (PLS) regression are used to model multiple submodels respectively according to size of subset. The weight coefficient of sub-model is obtained by Lagrange multiplier method, and the whole model is established by combining multiple sub-models. The simulation result shows that the proposed hybrid modeling method has a better predictive accuracy and robustness.