Scholarlink

<p>During the cutting process, both the chip color and temperature vary as the cutting tool wears. To improve tool life prediction, this study simultaneously monitors the temperatures of both the tool and the workpiece, and aims to develop and design temperature measurement equipment. The workpiece temperature will be measured using infrared rays, and the tool temperature will be measured using Type K thermocouples. The machining parameters are planned using the material removal rate (MRR). The experimental data will be optimized for the obtained chip chromaticity by using the K-means algorithm and the Ant Colony Optimization (ACO). Furthermore, Long Short-Term Memory (LSTM) neural networks are used for predicting tool life. After data optimization, the tool wear model of heavy-cutting (HC), medium-cutting (MC), and light-cutting (LC) improved by 12.67%, 1.71%, and 9.17%, respectively. Finally, the BPNN prediction results, the model with chip colorimetry, workpiece temperature, and tool temperature as input factors yields the best MAPE value. The tool wear of the HC, MC, and LC models is 4.71%, 4.20%, and 5.48%, respectively. Based on the LSTM modeling results, the tool wear of HC, MC, and LC models is methods show good prediction results in multi-feature prediction.</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

A study of the application of ant colony optimization in optimizing chip characteristics and cutting temperature for tool life prediction

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Shao-Hsien Chen,
  • Jyuan-Yu Chen

摘要

During the cutting process, both the chip color and temperature vary as the cutting tool wears. To improve tool life prediction, this study simultaneously monitors the temperatures of both the tool and the workpiece, and aims to develop and design temperature measurement equipment. The workpiece temperature will be measured using infrared rays, and the tool temperature will be measured using Type K thermocouples. The machining parameters are planned using the material removal rate (MRR). The experimental data will be optimized for the obtained chip chromaticity by using the K-means algorithm and the Ant Colony Optimization (ACO). Furthermore, Long Short-Term Memory (LSTM) neural networks are used for predicting tool life. After data optimization, the tool wear model of heavy-cutting (HC), medium-cutting (MC), and light-cutting (LC) improved by 12.67%, 1.71%, and 9.17%, respectively. Finally, the BPNN prediction results, the model with chip colorimetry, workpiece temperature, and tool temperature as input factors yields the best MAPE value. The tool wear of the HC, MC, and LC models is 4.71%, 4.20%, and 5.48%, respectively. Based on the LSTM modeling results, the tool wear of HC, MC, and LC models is methods show good prediction results in multi-feature prediction.