Convergence Study of Genetic Programming Approach to Estimate Surface Integrity on Additive Manufacturing Parts

Surface integrity in additive manufacturing technology is of utmost importance. In order to improve surface roughness of additive manufactured products a detailed understanding of the surface is required. One of the precise methods to model the surface integrity is Genetic Programming. However, convergence of the metaheuristic methods demands a thorough study of the approach. Throughout this paper, a new methodology is developed to improve the accuracy of the surface function extracted using Genetic Programming. The proposed approach introduces a deviation space to reduce the common deviational errors that inherently exists in acquiring the surface integrity using optical sensors. This space redefines the points in a new coordinate system which is computed using total least square method. To validate the process, actual case study on an additive manufactured part is examined for the convergence of the Genetic Programming on surface integrity.