Forecast Analysis Of Milling Vibration Signal Based On CEEMD And GWO-SVR

The splicing area of the cover will produce shock and vibration during processing,which will cause problems such as increased tool wear.Therefore,the establishment of a vibration prediction model for the splicing area of the splicing mold to predict the vibration and shock state during the splicing seam processing is of great significance for the selection of the actual processing preset parameters and the improvement of the processing quality.This paper takes the splicing mold of automobile outer cover parts as the research object,establishes the signal decomposition model and prospective prediction model of the splicing mold,and analyzes the influence of different prediction models on the vibration signal of the splicing area.Firstly,carry out the milling vibration signal acquisition experiment in the splicing area of the ball-end cutter milling curved surface mold,and conduct timedomain analysis of the milling vibration data under different processing parameters.Using the non-linear evaluation method based on Lyapunov exponent,the nonlinear characteristics of the vibration signal of the milling curved surface splicing die with the ball nose cutter are analyzed,and the nonlinear law of vibration of the milling curved surface splicing die is revealed.Secondly,in view of the modal aliasing problem of traditional empirical mode decomposition and the white noise interference of integrated empirical mode decomposition,the complementary integrated empirical mode decomposition method is used to overcome the shortcomings of empirical mode decomposition and integrated empirical mode decomposition methods.Establish a vibration signal decomposition model for the splicing area of the milling surface mold,calculate the correlation degree of the vibration signal,and determine the number of natural modal components used in the decomposition model.Thirdly,in view of the low prediction accuracy of traditional support vector machine regression and difficulty in training large sample data,a gray wolf pack optimization algorithm combined with a support vector machine regression prediction method is proposed,and the gray wolf pack algorithm is used to optimize the regression parameters of the support vector machine,and establish Prediction model based on gray wolf pack algorithm.Establish the support vector machine regression model based on genetic algorithm and the support vector machine regression model based on particle swarm algorithm and the gray wolf swarm algorithm prediction model for comparison experiments.Finally,a signal decomposition model based on complementary integrated empirical modes is used to decompose the experimental data,and the inherent modal components are input into each prediction model for prediction,and the prediction accuracy of different prediction models is compared.It proves that the vibration signal prediction model of the milling mold splicing area established in this paper has higher prediction accuracy,which is of great significance for the prediction research and actual processing of the milling mold splicing area.