Modelling the woven fabric strength using artificial neural network and Taguchi methodologies
Author:
- 29 February 2008
- journal article
- Published by Emerald Publishing in International Journal of Clothing Science and Technology
- Vol. 20 (2) , 104-118
Abstract
Purpose – Jacquard woven fabrics are widely used in various sections of upholstery industry, where mattress cover is one of them. Strength of jacquard woven mattress fabric depends on several factors. The objective of this study is to model the multi-linear relationship between fibre, yarn and fabric parameters on the strength of fabric using artificial neural network (ANN) and Taguchi design of experiment (TDOE) methodologies. Design/methodology/approach – TDOE was applied to determine the optimum design values and the contribution of each parameter. Robustness (performance) of models is measured by root mean squared error (RMSE). These tools will enable the user to predict the fabric strength from number of given inputs. It also provides the knowledge related to the contribution of fibre, yarn and fabric parameters on fabric strength. Fabrics tested in this study made from different fibre types and max/min level for several fabric and yarn-related parameters. The models generated with TDOE and ANN methodologies were compared with the actual experimental data. Findings – It was found that ANN model gives better approximation with the minimum RMSE. Research limitations/implications – The data taken from factory are related with jacquard woven fabric. Practical implications – This study has many practical implications that brings up a general approach for textile industry. During manufacturing, waste or scrap ratio can be reduced and production planning become more efficient. Originality/value – Firstly, before starting manufacturing in factory, we can easily predict the strength of woven fabric using the defined factors. This makes the model usable at the planning stage of the fabric. Secondly, the contribution of factors affecting fabric strength was determined. The ANN model generated in this study helps the engineers of planning department at the company easy to plan the manufacturing of fabric with a good estimation of fabric strength before the production order.Keywords
This publication has 20 references indexed in Scilit:
- Predicting the Pilling Propensity of Fabrics through Artificial Neural Network ModelingTextile Research Journal, 2005
- Predicting the fibre diameter of melt blown nonwovens: comparison of physical, statistical and artificial neural network modelsModelling and Simulation in Materials Science and Engineering, 2005
- A Process for Optimizing Sewing Conditions to Minimize Seam Pucker Using the Taguchi MethodTextile Research Journal, 2005
- Predicting Bursting Strength of Cotton Plain Knitted Fabrics Using Intelligent TechniquesTextile Research Journal, 2000
- An experimental study of needle heating in sewing heavy materials using infrared radiometryInternational Journal of Clothing Science and Technology, 1999
- Selection of optimal parameters in gas‐assisted injection moulding using a neural network model and the Taguchi methodInternational Journal of Quality Science, 1997
- An On-Line Fabric Classification Technique Using a Wavelet-Based Neural Network ApproachTextile Research Journal, 1996
- Yarn Strength Prediction Using Neural NetworksTextile Research Journal, 1995
- Theory and applications of neural networks for industrial control systemsIEEE Transactions on Industrial Electronics, 1992
- Quality Engineering Using Robust DesignTechnometrics, 1991