Scholarlink

<p>Milling chatter is a critical factor affecting machining quality. As a destructive phenomenon, it degrades workpiece surface finish, accelerates tool wear, and can hinder improvements in machining efficiency. Although extensive studies have been carried out on offline chatter prediction and online monitoring, reliable identification of early-stage chatter and extraction of physically interpretable chatter frequency information in complex machining environments remain challenging for online monitoring applications. To address these challenges, this study jointly analyzes the evolution of wavelet packet energy features and frequency consistency characteristics, and proposes an adaptive signal processing framework for milling chatter monitoring and frequency analysis. Specifically, a specific-band second-order differential wavelet packet energy entropy (SD-WPEE) method is developed to enhance the identification of the chatter incubation stage through sensitive feature extraction. In addition, statistical anomaly detection principles are incorporated to enable adaptive threshold updating, thereby reducing the influence of operating condition variations and background noise. For frequency analysis, a weighted least squares Burg (WLS-Burg) method with energy entropy–based weight optimization is employed to estimate the dominant chatter frequency, and frequency consistency is further evaluated using the frequency stability index (FSI). The proposed method is validated through multiple milling experiments under different machining conditions. The results demonstrate that the proposed framework can effectively characterize early-stage chatter energy evolution and analyze dominant chatter frequencies, indicating its potential applicability to machining stability analysis and chatter-related research.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on adaptive milling chatter monitoring method based on SD-WPEE and WLS-Burg

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Yadong Wu,
  • Hua Huang,
  • Xingyi Nie,
  • Zhijie Luo,
  • Min Nie

摘要

Milling chatter is a critical factor affecting machining quality. As a destructive phenomenon, it degrades workpiece surface finish, accelerates tool wear, and can hinder improvements in machining efficiency. Although extensive studies have been carried out on offline chatter prediction and online monitoring, reliable identification of early-stage chatter and extraction of physically interpretable chatter frequency information in complex machining environments remain challenging for online monitoring applications. To address these challenges, this study jointly analyzes the evolution of wavelet packet energy features and frequency consistency characteristics, and proposes an adaptive signal processing framework for milling chatter monitoring and frequency analysis. Specifically, a specific-band second-order differential wavelet packet energy entropy (SD-WPEE) method is developed to enhance the identification of the chatter incubation stage through sensitive feature extraction. In addition, statistical anomaly detection principles are incorporated to enable adaptive threshold updating, thereby reducing the influence of operating condition variations and background noise. For frequency analysis, a weighted least squares Burg (WLS-Burg) method with energy entropy–based weight optimization is employed to estimate the dominant chatter frequency, and frequency consistency is further evaluated using the frequency stability index (FSI). The proposed method is validated through multiple milling experiments under different machining conditions. The results demonstrate that the proposed framework can effectively characterize early-stage chatter energy evolution and analyze dominant chatter frequencies, indicating its potential applicability to machining stability analysis and chatter-related research.