引用本文:李鑫,史振宇,蒋森河,万熠,李欣.人工神经网络预测刀具磨损和切削力[J].控制理论与应用,2018,35(12):1731~1737.[点击复制]
LI Xin,SHI Zhen-yu,JIANG Shen-he,WAN Yi,LI Xin.Artificial neural network predicts tool wear and cutting force[J].Control Theory and Technology,2018,35(12):1731~1737.[点击复制]
人工神经网络预测刀具磨损和切削力
Artificial neural network predicts tool wear and cutting force
摘要点击 1911  全文点击 1101  投稿时间：2018-10-09  修订日期：2018-12-29
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2018.80769
  2018,35(12):1731-1737
中文关键词  机械加工，加工工具，切削力，材料磨损，神经网络，人工智能
英文关键词  Machining, Cutting tools, Cutting force, Wear of materials, Neural networks, Artificial intelligence
基金项目  国家自然科学基金重点项目；山东大学青年学者未来计划。
作者单位E-mail
李鑫 山东大学 316590272@qq.com 
史振宇* 山东大学 shizhenyu@sdu.edu.cn 
蒋森河 山东大学  
万熠 山东大学  
李欣 中国石化销售有限公司山东石油分公司  
中文摘要
      刀具磨损和切削力预测与控制是切削加工过程中需要考虑的重要问题。本文介绍了利用人工神经网络模型预测刀具磨损和切削力的步骤并且针对产生误差的因素进行分析。首先将切削速度、切削深度、切削时间、主轴转速和不同频带的能量值通过归一化法处理，作为输入特征值，对改进的神经网络模型进行训练。然后利用训练完成的神经网络模型预测刀具磨损和切削力。结果表明：神经网络模型能够综合考虑加工过程中更多的影响因素, 与经验公式结果对比，具有更高的预测精度。研究结果表明神经网络模型预测刀具磨损和切削力具有可行性和准确性，为刀具结构的优化及加工参数的选择提供了依据。
英文摘要
      Tool wear and cutting force prediction and control are important problems to be considered in the machining process. In this paper, the process of predicting tool wear and cutting force by using artificial neural network model are introduced, and the factors that produce error are analyzed. Firstly, the cutting speed, cutting depth, cutting time, spindle speed and energy value of different frequency bands are treated by normalization method, which are used as input eigenvalues, and the improved neural network model is trained by those input eigenvalues. Then the tool wear and cutting force are predicted by using the trained neural network model. The results show that the neural network model can consider more factors in the machining process, and compared with the results of empirical formulas, it has higher prediction accuracy. The results show that the neural network model has the feasibility and accuracy in predicting tool wear and cutting force, and can be helpful for the optimization of tool structure and the selection of machining parameters.