Distorsion Prediction of Additive Manufacturing Process using Machine Learning Methods

Abstract

Additive Manufacturing is a widely used technology; however, it also has several open questions. In the modelling phase, it is necessary to predict undesired distortions. There are several finite-element based simulation tools for this purpose, but these are costly and resource-intensive. This paper presents a novel approach based on several Machine Learning methods (decision trees, random forest, gradient boosted trees, support vector machines, deep learning) to speed-up this process. The results show that it is possible to give accurate predictions with these methods.