Abstract

This study explores the development and application of data-driven control techniques for managing the power of a laser system used in the in-situ consolidation process of thermoplastic materials (ISC). We discuss the correlation among the main variables - temperature, power, layer number, and lamination speed - and how these interactions inform the design of our control models. Two types of prediction models, multiple polynomial regression and support vector machines are compared. Though the software solution developed here is for testing purposes and not for production, we demonstrate the utility and flexibility of machine learning control approaches for this type of manufacturing process.