A multi-objective parameter optimization and decision-making method for multi-pass end milling with firefly algorithm and Markov clustering
Xu-Lin Cai,Wen-An Yang,Xue-Feng Yang,You-Peng You
DOI: https://doi.org/10.1007/s40430-024-04740-1
IF: 2.361
2024-03-12
Journal of the Brazilian Society of Mechanical Sciences and Engineering
Abstract:Machining parameter optimization holds profound significance within the domain of milling operations, as the selection of these parameters can exert a substantial influence on both production efficiency and the quality of produced components. Notably, the integration of the chatter stability constraint into the optimization model remains underrepresented in the existing literature, despite its pivotal role in ensuring tool safety and machining quality. Furthermore, while several multi-objective optimization (MOO) algorithms have been devised, their effectiveness is often compromised due to the complexity of the milling parameter optimization model. In response to these challenges, this study developed an updated full discretization method (UFDM)-based three-dimensional (3-D) stability prediction model for multi-pass end milling. A milling parameter optimization model is meticulously formulated herein, which simultaneously optimizes the number of passes, feed rate, spindle speed, axial cutting depth, and radial cutting depth in multi-pass end milling while complying with constraints including machine tool performance, tool life, workpiece characteristics, and 3-D stability, to minimize production time and surface roughness. Moreover, a novel MOO and decision-making system for milling parameters is developed to solve the constructed model and assist decision-making, which includes a novel Markov clustering (MCL)-enabled MOO firefly algorithm (MEMOFA), a firefly algorithm-enabled MCL and a pseudo-weight-coefficient-vector-based decision-making method. The empirical findings encompassing six benchmark functions unequivocally attest to the superior performance of the elaborated MEMOFA in matters of convergence, diversity, and spread. Furthermore, an extensive optimization for the milling parameters model is also carried out to verify the superiority of the developed MOO and decision-making system in solving milling parameters optimization.
engineering, mechanical