Optimisation decision of machining process parameters considering milling energy consumption and specific cutting energy
Yang Xie, Shulong Mei, Chaoyong Zhang
Abstract
To address the challenges of high energy consumption and specific cutting energy in CNC milling operations, this study proposes a predictive–optimization integrated framework for multi-objective process parameter optimization. A high-fidelity energy-efficiency prediction model is first established by combining the Whale Optimization Algorithm (WOA) with a Backpropagation Neural Network (BP), enabling accurate estimation of milling power and specific cutting energy based on spindle speed, feed rate, depth of cut, and milling width. On this basis, a multi-objective optimization scheme is developed using the second-generation Non-dominated Sorting Genetic Algorithm (NSGA-II), targeting the simultaneous minimization of the two energy indicators. The entropy-weighted TOPSIS method is subsequently employed to rank the Pareto-optimal solutions and determine the most favorable parameter combination. Experimental validation demonstrates that the optimized parameters result in reductions of 12.33 % in milling power and 20.01 % in specific cutting energy compared to empirical settings. The proposed method not only realizes the synergistic optimization of energy-related objectives but also demonstrates the effectiveness and robustness of WOA-BP and NSGA-II in addressing complex process optimization problems, offering a promising approach to support energy-efficient and sustainable CNC milling.