Abstract:The chatter occurrence in the milling process is a key factor that limits the machining efficiency and quality. The milling stability is mainly dependent on the process parameters and the dynamic characteristics of the tool-workpiece system. However, the system dynamics vary with the changes of the machining position and tool properties. Considering these multiple influence factors, a method is proposed to predict the milling stability and determine optimal machining parameters based on the bootstrap aggregating (Bagging) and non-dominated sorting genetic algorithm-II (NSGA-II). First, the orthogonal experiment design is used to divide the working space of the machine tool into different machining positions. Under each position, the impact testing is carried out at the tool tip for different tool overhang lengths to obtain the corresponding frequency response functions (FRFs). Then, the limiting axial cutting depth aplim values are theoretically predicted using the tool tip FRFs and machining parameters. With the sample information, the bagging algorithm is taken to establish a model of predicting the aplim, where the inputs are the displacements of the moving parts (x, y, z), tool diameter (d), tool overhang length (h), spindle speed (n), cutting width (ae) and feed rate per tooth (fz). Taking these process parameters {x, y, z, d, h, n, ap, ae, fz}as the design variables, a multi-objective optimization model is constructed to balance the machining efficiency and tool life. Additionally, the pre-established aplim predication model is utilized to express the milling stability constraint. Then, the multi-objective optimization model is solved by the NSGA-II and the Pareto-optimal set is obtained. The entropy weight method and the technique for order preference by similarity to an ideal solution (TOPSIS) are combined to select a unique optimal solution from the Pareto-optimal set. A three vertical machining center was taken to carry out a case study. The prediction accuracy of the established Bagging model of aplim was 2.99%, and no chatter was observed when performing the milling test with the finally determined optimal process parameters. These experimental results validated that the feasibility of the proposed method for predicting the milling stability and selecting optimal process parameters under multiple influence factors.