The geometric predictive model of properties for systems with interval parameters

Abstract

We consider interval geometric modeling of complex multi parametric systems having a set of parameters of different character. Some of the parameters may have interval indefiniteness. System information basis is incomplete one. The processed information depends on continuous, discrete and conventional data. Geometric model has a form of matrix and each element of it corresponds to some state of the system. Each state is described by interval function of continuous input and output parameters. The set of interval functions generates some discrete set of multidimensional surfaces in discrete space. We use this approach and our modeling algorithm to find predictive model of drape coefficient. The algorithm is based on linear approximation of numerical factors in factor spaces. Interval functions make it possible for us to vary some numerical factors within the given intervals. As an example, the interval model of fabric drape coefficient is found. Fabric thickness and closeness of texture are considered as input parameters.

Keywords

• geometric model
• interval analyses
• prediction
• fabrics
• drape coefficient