Development of an adaptive, self-learning control concept for an additive manufacturing process

Abstract

Error avoidance in high-precision manufacturing processes becomes more important for numerous state-of-the-art technologies. Selective laser melting is one of these technologies offering large potentials in the production of complex and flexible metal products. As the technology is relatively new, it is vulnerable for errors, given that the process parameters are not measured yet. A novel multilevel control concept, incorporating several sensors, has the potential to reduce errors significantly. For inner cascade control, the laser power will be adjusted by measurements with an intensity sensor for wavelengths in the visible range. This sensor is integrated into the optical path of the laser beam. An adapted self-learning strategy supports the stability of the process by updating the parameters of the used multidimensional model in order to attenuate environmental influences or shifts within the process. This work presents the concept of the control approach, first measurement results and the required relations between measurement, process and control parameters.