Flexible process planning based on predictive models for machining time and energy consumption

Machining time and energy consumption are common process planning evaluation metrics. However, it is difficult to obtain accurate machining time and energy consumption values in the process planning stage, which may result in non-optimal obtaining process routes. To solve these problems, this paper proposes a flexible process planning method based on processing time and energy consumption prediction models. The historical processing data is analyzed, and a prediction model of machining time and energy consumption based on a radial basis function neural network is proposed. Considering the flexibility of multiple processes, a mathematical model of the flexible process planning problem based on the machining time and energy consumption prediction model is established. A multi-objective algorithm using a multidimensional real number coding method is proposed for problem-solving with the completion time and energy consumption minimization as the optimization objective. Two algorithms with better multi-objective optimization performance, the strength Pareto evolution algorithm 2, and the dominance ranking genetic algorithm II are selected for comparative analysis to obtain a better effect of optimization. The better algorithm is selected with the best parameters. Finally, case studies are conducted to verify the validity of the prediction model and the flexible process planning model.