Discrete physics as a medium for surrogate models: the LPBF case

The role of digital twins in real-time optimizing LPBF is crucial. To achieve, however, the implementation of digital twins, there is need for surrogate models. This will facilitate the intuitiveness of the models, but will also drive the link between quality assessment and process control. Then, there are requirements for knowledge on the evolution of the phenomena in space and time. To this end, herein, ARX models are considered as potential solutions and a link to new computational methods is proposed. The benefits of such surrogate models are presented; they are fast, adaptive and intuitive, offering cognitive capabilities