Optimized Milling Processes for Carbon Emissions Based on Hybrid Multi-Objective Algorithms

To address the accelerated implementation of global carbon neutrality policies, this study proposes a milling process parameter control system that is predictive, optimizable, and low-carbon oriented. The system employs a non-destructive clamp-type ammeter to measure the instantaneous current of the machine tool in real time during machining, enabling accurate estimation of energy consumption and carbon emissions, and providing high-resolution data for subsequent modeling. Based on these data, a continuous prediction model is constructed using multiple polynomial regression combined with cubic spline interpolation. This model is then integrated into a hybrid multi-objective optimization framework that combines Multi-Objective Particle Swarm Optimization (MOPSO) and the Nondominated Sorting Genetic Algorithm III (NSGA-III) to optimize carbon emissions, machining time, and surface roughness. Experimental results demonstrate that, under both fixed cutting depth and full decision space scenarios, the maximum carbon emission reduction rates reached 19.12% and 28.49%, respectively, while the best verification error rates in actual machining were −0.66% and −6.82%. Furthermore, an MVC architecture, combined with an MSSQL database and the MQTT protocol, was employed to develop a responsive web interface, enabling real-time data transmission, visualization, and interactive decision support.