Линейные и квадратичные нечеткие иерархические окрестностные модели процесса производства холоднокатаной стали

Abstract

The development of a simulation model for the production process of cold-rolled steel is relevant, as it allows the prediction of mechanical properties that are formed after annealing, without conducting a full-scale experiment. The aim of the work is to apply a neighborhood approach to modeling the process under consideration based on linear and quadratic dynamic neighborhood models, and to compare the simulation results with production data. The definition of dynamic neighborhood models with a fuzzy hierarchical structure, distinguished by fuzzy two-level neighborhood connections between nodes of the first and second level of the model is given. Constructed linear and quadratic clear non-hierarchical and fuzzy hierarchical dynamic neighborhood models to calculate the yield strength, temporal resistance and relative elongation. The analysis of model and production data is performed on the basis of relative identification and modeling errors for the considered models.