SIMULATION MODEL FOR PREDICTIVE EVALUATION OF POST-PRINTING PROCESS QUALITY

Abstract

The methodology for predictive evaluation of the quality of post-press processes has been developed using fuzzy logic methods and tools. A fuzzy system for assessing the quality of the investigated technological process has been created, consisting of two main components: fuzzification and defuzzification. Fuzzification converts crisp values into fuzzy variables. The stages of fuzzification include generating partial quality indicators, defining a universal set of values and corresponding term sets, developing a fuzzy inference model, processing membership functions of linguistic variables, creating knowledge bases, knowledge matrices, and fuzzy logic equations. Defuzzification converts fuzzy results into solutions applicable in real-world conditions. This process involves creating tables of membership function values, calculating fuzzy logic equations, and obtaining a specific integral quality indicator for post-press processing of book publications.

An algorithm for a simulation model for predictive evaluation of post-press process quality has been developed based on the fuzzy system. The software product Post Press Forecast was created to determine the predicted integral quality indicator based on user-defined values of universal sets of linguistic variables. The reverse operation involves selecting the desired integral indicator value and setting acceptable deviation limits, resulting in a table of possible parameter combinations for factors. This enables the selection of optimal parameters to achieve the expected result, significantly reducing the number of substandard copies and shortening the time-to-market for book publications. As a result, production profitability, quality, and competitiveness of the final product are enhanced.