Scholarlink

<p>Computer Numerical Control (CNC) machines play an essential part in industrial production to ensure efficiency and reliability. Developing a useful diagnosis system is necessary to achieve zero-defect production during the machining processes. However, detecting anomalies in CNC machines significantly avoids catastrophic failures and leads the industry to maintain efficient operation. Various techniques utilized in this field face several limitations, such as insufficient labeled data, computational complexity, varying degrees of noise effects, and high maintenance costs. To overcome the above challenges, a model named Distributed Gradient AdaBoost Convolutional Network (DiGACN) is proposed in this research for detecting anomalies present in the machine tools by eliminating the drawbacks of existing methods. Integration of AdaBoost and the Gradient Boosting Machine (GBM) algorithm improves the DiGACN model’s performance accuracy by reducing the overfitting issues. Moreover, the probability-based adaptive scoring is exploited to adjust the behaviour of the model based on input data, and it enables minimizing the error value. The DiGACN model attained 97.93% specificity, 97.91% sensitivity, and 97.92% accuracy, with 90% of training data using the CNC Machining Dataset, respectively.</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

DiGACN: Distributed Convolutional Network for Anomaly Detection in Computer Numerical Control Machines

  • 期刊论文

  • 发表日期：

    2026年7月3日

查看全文
  • Y. Mohana Roopa,
  • K. Sree Latha,
  • B. Bhaskar Reddy,
  • M. Ramesh Babu,
  • Kakarla Hari Kishore

摘要

Computer Numerical Control (CNC) machines play an essential part in industrial production to ensure efficiency and reliability. Developing a useful diagnosis system is necessary to achieve zero-defect production during the machining processes. However, detecting anomalies in CNC machines significantly avoids catastrophic failures and leads the industry to maintain efficient operation. Various techniques utilized in this field face several limitations, such as insufficient labeled data, computational complexity, varying degrees of noise effects, and high maintenance costs. To overcome the above challenges, a model named Distributed Gradient AdaBoost Convolutional Network (DiGACN) is proposed in this research for detecting anomalies present in the machine tools by eliminating the drawbacks of existing methods. Integration of AdaBoost and the Gradient Boosting Machine (GBM) algorithm improves the DiGACN model’s performance accuracy by reducing the overfitting issues. Moreover, the probability-based adaptive scoring is exploited to adjust the behaviour of the model based on input data, and it enables minimizing the error value. The DiGACN model attained 97.93% specificity, 97.91% sensitivity, and 97.92% accuracy, with 90% of training data using the CNC Machining Dataset, respectively.