Application of Mating Behaviour-based Metaheuristic Algorithms for Optimizing Green Machining Processes

Abstract

In the modern-day manufacturing landscape, the conventional material removal processes are being replaced by green and environment-friendly processes to enhance occupational safety, health standards, and reduce energy consumption and carbon footprints. They can practically support sustainable manufacturing, particularly in reducing use of lubricants and energy, thereby protecting the environment from various perilous effects. This paper optimizes two green machining processes (dry milling and hard turning) using five mating behaviour-based metaheuristics, i.e. firefly mating algorithm (FMA), naked mole-rat algorithm (NMR), butterfly mating optimization algorithm (BOA), barnacle mating optimization algorithm (BMOA) and American zebra optimization algorithm (AZOA). Experimentally validated response surface methodology-based models reported in the literature are adopted as the objective functions and all the algorithms are implemented under identical computational settings to ensure fair comparison. Their efficacy is contrasted in regard of solution accuracy and variability, and time complexity. In multi-objective optimization, the Pareto fronts are also developed to identify the optimal intermixes of the process parameters. Their algorithmic performance is further validated using hypervolume, spacing, Friedman's mean rank test and Wilcoxon rank-sum test. For both the processes, FMA supersedes its competitors in attaining the optimum mixtures of the corresponding machining parameters. During dry milling of stainless steel 304, the single- and multi-objective optimization results show that FMA achieves 28.57%, 25% and 13.04%, and 3.57%, 5.56% and 7.05% improvements in power factor, surface roughness and energy consumption, respectively, in comparison to the prior observations. In hard turning of AISI D3 steel, the single- and multi-objective optimization solutions exhibit those improvements as 31.07%, 16.45% and 8.03%, and 18.55%, 7.22% and 2.92% in surface roughness, power consumption and flank wear, respectively. Thus, FMA appears as an effective and robust optimization tool in identifying near-global optimal machining conditions, providing practical decision support for energy-efficient, high-quality and sustainable manufacturing.