A Fuzzy Criterion for Selecting Relevant Process Parameters for the Development of Nonwoven Products

In this paper, we present a new fuzzy selection criterion which takes into account both the sensitivity of measured data and human knowledge concerning the relations between process parameters and quality features of nonwoven product. This selection criterion permits to rank the process parameters of the related nonwoven production line and take the most relevant ones as input variables of the model for designing new nonwovens products. Compared with the classical selection criteria, the proposed method is more robust and less sensitive to proximities of measurement and uncertainties. This method has been validated using real data collected from a nonwoven process and a predefined questionnaire filled by experts.