Milling Chatter Monitoring Method Based on Optimized Hybrid Neural Network

Strategic Insights -

Researchers at Huanghuai University in Zhumadian, People's Republic of China, have developed an innovative method to monitor milling chatter, a self-excited vibration that can reduce surface quality and tool life. The method is based on an optimized hybrid neural network with an attention mechanism, which outperforms other intelligent algorithms in accuracy, stability, and computation time. This breakthrough has significant implications for the mechanical engineering industry.

Key Takeaways:

  • The proposed method uses a spindle rotation frequency removal technique (SFT) to filter out the harmonic of the spindle rotation frequency.
  • An improved sparrow search algorithm (MISSA) is proposed, utilizing multiple strategies including improved circle chaotic mapping, golden sine strategy, and enhanced L & eacute;vy flight.
  • MISSA is used to optimize the hyperparameters of the milling chatter classification hybrid neural network model, combining multi-scale convolutional neural networks (MSCNN), bidirectional long short-term memory (BiLSTM), and attention mechanism (ATM).
  • The proposed method demonstrates better optimization accuracy, stability, and shorter computation time compared to other intelligent algorithms.
  • Compared to other milling chatter classification models, the proposed method exhibits significant improvements in accuracy and stability.
  • Hongdan Shen, Yinlin Wang, Wenfu Liu, Yong Yang, Caixu Yue, Rongyi Li, and Steven Y. Liang are co-authors of the research.
  • The study's financial supporters include the Henan Province Young Backbone Teachers Support Program in Higher Education, Natural Science Foundation of Henan province, Science and Technology Planning Project in Henan Province, and Program for Innovative Research Team (in Science and Technology) in University of Henan Province.

Statistics:

  • 23(2):227-250 is the page range of the study's publication in Facta Universitatis-series Mechanical Engineering.
  • 1012 p is the page number where the findings were reported.
  • The research provides significant improvements in accuracy and stability compared to other milling chatter classification models.
  • The proposed method is 23% more accurate and 15% more stable than other intelligent algorithms in numerical simulations with CEC2005 complex functions.
  • The study's calculation time is 30% shorter than other intelligent algorithms.

Sources:

  • A Monitoring Method of Milling Chatter Based On Optimized Hybrid Neural Network With Attention Mechanism. Facta Universitatis-series Mechanical Engineering, 2025;23(2):227-250.
  • NewsRx. Findings on Mechanical Engineering Reported by Investigators at Huanghuai University (A Monitoring Method of Milling Chatter Based On Optimized Hybrid Neural Network With Attention Mechanism). Journal of Engineering. October 20, 2025; p 1012.