A0351
Title: Nonlinear models for mixture experiments including process variables
Authors: Shroug Alzahrani - University of Southampton (United Kingdom) [presenting]
Abstract: Mixture experiments are applied in a variety of fields, for example, food processing, chemical engineering, and product quality improvement, that mix multiple components to perform. The measured response in such experiments is a function not of the amount of the mixture components but their proportions. In some mixture experiments, the blending properties of the mixture may be affected by the processing conditions, such as temperature and pressure, which adds a layer of complexity to the modeling. Such mixture experiments are known as mixture-process variables experiments. For example, while the flavour of a cake depends on the proportions of the cake ingredients, process variables such as cooking time and cooking temperature affect the taste of the cake as well. The experimental region is constrained naturally, as each proportion of mixture components must be greater than zero, and the proportions of all mixture components must sum to one. Often, there are additional restrictions on the proportions when lower and upper limits bound them. A new class of nonlinear models is proposed for mixture-process variables experiments and is compared with standard models from the literature. Moreover, extended forms for modified fractional polynomial models are suggested to fit data from mixture-process variables experiments.
