Development of models for forecasting and classification of a printing quality of a low cost 3D printer

Abstract

At the present, the actual task is using 3D printers for the manufacture of certain objects with a given level of price / quality ratio. In many cases, it is economically feasible to use a low cost 3D printer. Therefore, it is necessary to have models that predict and classify the printing quality of such printers. The work has involved the development and assembly of a low cost 3D printer. For this purpose, the creation of geometric models of the component parts and the printer itself was carried out, and engineering calculations and optimization of the received designs were performed. It has been developed a printer control system. An experiment was conducted to produce cubes with different printing parameters on such printer. Based on regression analysis, linear and logistic regressions were constructed. Linear regression will allow to assess the quality level of the result depending on the printing parameters, and the logistic regression will allow to classify and predict the probability of manufacturing objects with a given quality level. The influence of each of the print parameters on the quality and result of the classification was analyzed.