Optimization of milling parameters based on GA-BP neural network

Since the selection of machining parameters usually relies on the experience of workers or traditional calculation formulas, the traditional prediction method has the disadvantages of a complicated arithmetic process, large prediction deviation, and high consumption cost, making it challenging to meet the increasing demand of production and processing. Therefore, this paper proposes a machining quality prediction model based on the GA-BP neural network. Through experiments, it verifies the data-fitting ability of the prediction model and then takes the prediction model as the optimization objective, culminating in a multi-objective optimization model for process parameters based on the NSGA-II algorithm. Experiments demonstrate that the cutting force and surface roughness obtained by the optimization model are 3.6% and 10.0% lower than those obtained by the empirical parameters, respectively, leading to reductions of 3.6% and 10.6%, which verified the optimization effect of the model.