A Parameter-Less Multi-Objective Optimization Framework for Additive, Thermal, and Subtractive Manufacturing Processes

Multi-objective optimization has become an indispensable tool for solving engineering design and manufacturing problems involving multiple conflicting objectives. This paper presents a novel parameter-less multi-objective optimization (MOO) framework that combines the strengths of evolutionary MOO techniques with the parameter-free search philosophy of the Jaya and Rao algorithms. The proposed framework incorporates non-dominated sorting, elite archiving, and crowding-distance mechanisms to achieve an effective balance between convergence and diversity while eliminating the need for algorithm-specific control parameters. The proposed framework is first validated on sixteen widely used unconstrained benchmark problems comprising five ZDT, seven DTLZ, two IDTLZ, and two SDTLZ test suites using the maximum number of function evaluations reported in the literature. Its performance is evaluated using five widely accepted quality indicators, namely Generational Distance (GD), Inverted Generational Distance (IGD), Hypervolume (HV), Spacing (SP), and Spread (SD). The benchmark results demonstrate that the proposed framework produces competitive Pareto-optimal fronts and exhibits excellent convergence, diversity, and solution distribution compared with several state-of-the-art evolutionary multi-objective optimization algorithms. The practical applicability of the proposed framework is demonstrated through five representative manufacturing optimization problems involving Selective Laser Melting, Microwave Hybrid Heating, Sustainable Machining, Wire Electrical Discharge Machining, and Wire Arc Additive Manufacturing. These case studies encompass additive, thermal, subtractive, and many-objective manufacturing optimization problems with conflicting performance measures. The generated Pareto-optimal solutions are subsequently ranked using the recently developed BHARAT (Best Holistic Adaptable Ranking of Attributes Technique) multi-attribute decision-making method to identify the most suitable compromise solutions. The obtained results demonstrate that the proposed parameter-less MOO framework provides a simple approach with competitive convergence, diversity, and decision-support capabilities for complex manufacturing optimization problems.