An uncertainty approach for optimization of production parameters-a case study in an extrusion molding process

作者：Alberto Rodriguez Picon Luis

来源：International Journal of Advanced Manufacturing Technology, 2017, 90(1-4): 167-176.

DOI：10.1007/s00170-016-9358-x

• Uncertainty
• Optimization
• Normal distribution
• Production parameters

This paper aims to provide an approach in the optimization of the functional requirements and design parameters of a production process based on uncertainty. Such approach is developed considering that the uncertainty arises from a manifestation of information deficiency. In this case, the uncertainty is characterized by the inability of the production system to completely satisfy the corresponding functional requirements reflected in the product. All production processes generate information, either directly or indirectly, from its operations; the characteristics of this information can be viewed in terms of uncertainty to analyze its behavior with respect to the variation of the functional requirements and the process parameters. The uncertainty approach is defined considering that the behavior of the functional requirements and process parameters follow a normal distribution; from this distribution, the uncertainty is obtained via the continuous form of the Shannon entropy. Regression modeling is used as a tool to relate the functional requirements of the process to their corresponding parameters, such that the multiple regression model can be established and extended as a time function to define a model to deal with uncertainty. In addition, control parameters are obtained by using step-wise-regression. This modeling is presented for assessment, optimization, and control purposes of an extrusion molding process.

• 出版日期2017-4