Research On Milling Chatter Recognition Based On Multi-sensor Information Fusion

Chatter is a complex non-linear,non-stationary self-excited vibration that occurs in milling,turning,drilling,grinding,and other machining processes.In the milling process of thin-walled parts or high-strength and high-hardness materials,cutting chatter is more likely to occur.Cutting chatter can reduce the quality and efficiency of machining,as well as the working life of cutting tools and machine tools.Therefore,online monitoring and early recognition of milling chatter are of great significance.In this paper,in-depth research has been conducted on milling chatter recognition,and the main work is summarized as follows:The milling stability is analyzed and milling experiments of multi-sensor fusion are designed.Based on the principle of regenerative chatter,a milling dynamic model is established,and the milling stability region is solved using the analytic method.The natural frequency,damping ratio and stiffness of the milling system are obtained by modal hammer experiments,and a stability lobes diagram is drawn to analyze the impact of various modal parameters on milling stability.According to the stability lobes diagram,the multi-sensor milling monitoring experiments integrating force sensors and acceleration sensors are designed,and the milling signal provides data support for chatter recognition later.To optimize the key parameters of variational mode decomposition(VMD)and support vector machine(SVM),an improved sparrow search algorithm based on multi-strategy fusion is proposed.In the process of optimizing key parameters of VMD and SVM,aiming at the shortcomings of sparrow search algorithm in population initialization and sparrow position update,the improved Circle chaotic map,sine cosine algorithm,Lévy flight strategy,and adaptive t distribution are used to improve it.The simulation experiments,Wilcoxon rank sum test and time complexity analysis have verified the advantages of the improved algorithm in terms of search ability,convergence speed,optimization accuracy and stability,providing algorithm support for the optimization of key parameters.The variational mode decomposition optimized by the improved sparrow search algorithm is used to analyze milling signals,and the multi-scale permutation entropy of the decomposed sensitive components is extracted as chatter recognition features.According to the principle of minimum average envelope entropy,the improved sparrow search algorithm is used to optimize the number of modes and penalty parameters of variational mode decomposition.The optimized variational mode decomposition is used to analyze the vibration and force signals of milling experiments,and the decomposed sensitive components are selected based on the correlation coefficient method.The multi-scale permutation entropy of the sensitive components is extracted as the chatter recognition feature.A milling chatter recognition model is established using support vector machine optimized by the improved sparrow search algorithm.The multi-scale permutation entropy of the milling experimental vibration signal and force signal is used as the model input,and based on the principle of the highest average recognition accuracy of the model using the 5-fold cross validation method,the improved sparrow search algorithm is used to optimize the penalty parameters and kernel function parameters of the support vector machine,and a milling chatter recognition model based on multi-sensor information fusion is established,which can identify chatter in time during chatter incubation stage,so as to ensure stable milling and improve milling efficiency.