Factorial experimental design and the investigation of various regression models for generating optimal empirical formulas.

Abstract

One of the most important challenges of today’s industry is to be able to optimize various manufacturing processes as accurately, reliably and cost-effectively as possible. The market demands set by customers are constantly changing: the production cycle and development pace of a product are getting shorter, their variations are spread over a wide spectrum, to which, if the manufacturer is unable to react in time, it can easily fall behind its competitors in the market competition. Sustainability expectations also place serious expectations. To meet these challenges, continuous monitoring and analysis of manufacturing processes and their parameters is essential. For this, it is advisable to use mathematical models that can help us predict the impact of a given parameter combination on the desired result, for example, surface roughness in machine manufacturing technology. In this study, we present the method of factorial experimental design and different regression models using the results of a previously conducted series of experiments, and then perform a comparative analysis to determine which model best approximates the measured results.