A multi-stage adaptive genetic algorithm for multi-objective optimization of machining parameters in complex part manufacturing

This paper proposes a multi-stage adaptive genetic algorithm for multi-objective optimization of machining parameters in complex structural part manufacturing. The model jointly considers machining accuracy, processing time, and energy consumption under parameter coupling and nonlinear constraints. An adaptive crossover-mutation strategy and a multi-stage evolutionary mechanism are introduced to improve convergence stability and balance exploration with exploitation. Experiments are conducted on a self-developed Python-based simulation platform using 12,400 samples covering 30 part structures, 12 machining processes, and 36 parameter variables. Compared with the standard GA and NSGA-II, the proposed method reduces the average machining cycle to 13.1 h and the order delay rate to 3.8%, while maintaining a resource utilization rate of 87.4%, demonstrating stable performance in complex machining scenarios.