Machine Learning Based Heuristic Technique for Multi-response Machining Process

Abstract

Manufacturing process variables influence the quality of products substantially. It is unquestionably difficult to model the manufacturing processes that include a large number of variables and responses. Development of the multi-objective surrogate models for the manufacturing processes could be computationally and economically expensive. In this article, a generic multi-objective surrogate-coupled heuristic algorithm is employed that needs small amount of experimental data as input, and predicts precise responses with quick Pareto solutions. The proposed algorithm is verified with different cases collected from the literature based on the CNC turning, centerless cylindrical grinding, and micro milling machining and shown to produce some interesting results.