In many industrial situations, variation in naturally occurring import material quality results in inconsistency in the end product quality. Since such raw materials are from natural sources, the variation in their quality cannot be controlled. Several methods, such as statistical process control (SPC), engineering process control (EPC), Taguchi's robust process design, etc. are available for controlling the unwanted variability in production processes. Another method is to sort the raw materials into batches of similar quality which need similar processing conditions. The purpose of this article is to develop a methodology for such sorting. The methodology is investigated with a focus on convergence properties of the numerical algorithm and sensitivity to different utilizations. It is also tested for a model with more than one variable descripting raw material properties and is extended by inclusion of a penalty function that penalizes unfavorable process settings. (24 refs.)
展开▼