Scholarlink

<p>Machine tools play a crucial role in the manufacturing industry. In response to the poor generalization of traditional machine tool energy consumption prediction, this paper proposes a refined staged energy prediction method of computer numerical control (CNC) milling process based on whale optimization algorithm (WOA) optimized extreme gradient boosting (XGBoost). By dividing the machining process into four stages and utilizing data processing and feature selection, the WOA algorithm is used to optimize the hyperparameters of the XGBoost algorithm, establishing power and time models for each stage, and then calculating energy consumption. On this basis, experimental research was conducted to predict milling energy consumption using the proposed method, and compared with actual energy consumption and three other algorithm models. The verification result shows that the accuracy of the method is 95.49%, which validates the effectiveness of the method. Compared with traditional methods, the improved staged energy prediction accuracy has increased by 0.177%, the minimum accuracy value has increased by 4.477%, and the standard deviation of accuracy has decreased by 19.274%. The improved method exhibits smaller and more stable accuracy fluctuations, along with stronger predictive ability, providing more accurate energy model support for milling energy optimization.</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

Refined staged energy prediction method of CNC milling process based on WOA Algorithm with XGBoost hyperparameters optimization

  • 期刊论文

  • 发表日期：

    2025年11月15日

查看全文
  • Shun Jia,
  • Bingying Wu,
  • Shang Wang,
  • Biao Kong,
  • Shuyu Li,
  • Guoqiang Zhang,
  • Jinyun Song,
  • Anbang Li,
  • Yang Sui,
  • Zhaojun Steven Li

摘要

Machine tools play a crucial role in the manufacturing industry. In response to the poor generalization of traditional machine tool energy consumption prediction, this paper proposes a refined staged energy prediction method of computer numerical control (CNC) milling process based on whale optimization algorithm (WOA) optimized extreme gradient boosting (XGBoost). By dividing the machining process into four stages and utilizing data processing and feature selection, the WOA algorithm is used to optimize the hyperparameters of the XGBoost algorithm, establishing power and time models for each stage, and then calculating energy consumption. On this basis, experimental research was conducted to predict milling energy consumption using the proposed method, and compared with actual energy consumption and three other algorithm models. The verification result shows that the accuracy of the method is 95.49%, which validates the effectiveness of the method. Compared with traditional methods, the improved staged energy prediction accuracy has increased by 0.177%, the minimum accuracy value has increased by 4.477%, and the standard deviation of accuracy has decreased by 19.274%. The improved method exhibits smaller and more stable accuracy fluctuations, along with stronger predictive ability, providing more accurate energy model support for milling energy optimization.