基于循环神经网络的切削力无传感器监测

doi:10.3901/JME.2025.06.014

摘要/Abstract

摘要： 切削力对切削状态的变化具有高度敏感性和快速响应能力，是用于加工状态监测和自适应加工最有价值的物理量。由于无须引入额外的传感元件，基于数控系统内部伺服监测信号的切削力在线预测方案，有潜力实现切削力的长期低成本准确监测。然而，伺服监测信号和切削力之间的关系十分复杂，为此提出基于循环神经网络的切削力在线预测方法。首先将基于机床伺服监测信号的进给轴切削力预测问题定义为自适应时滞的非线性动态系统建模问题，进而引入长短期记忆(Long short-term memory,LSTM)神经网络和门控循环单元(Gated recurrent unit,GRU)神经网络两类循环神经网络直接从端到端观测数据中学习动态预测模型。通过一组变转速变进给孔铣削试验构造时变工况下的切削激励数据集进行对比验证，结果表明，对于动态特性更为复杂的X轴，LSTM神经网络的预测性能更好，相对方均根误差为17.62%；对于动态特性相对简单的Y轴，GRU神经网络的预测性能更好，相对方均根误差为11.74%。

关键词: 切削力, 加工状态监测, 自适应加工, 循环神经网络, 伺服监测信号

Abstract: Cutting force is highly sensitive and capable of rapid response to changes in cutting state, which is considered as the most valuable physical quantity for machining state monitoring and adaptive machining. Since there is no need to introduce additional sensing components, an online prediction cutting force solution based on inherent servo monitoring signals in CNC systems has the potential to achieve long-term, low-cost, and accurate monitoring of cutting force. However, the relationship between servo monitoring signals and cutting force is very complex. Therefore, a cutting force online prediction method based on recurrent neural networks is proposed. Firstly, the problem of feed-axis cutting force prediction based on machine tool servo signals is defined as a nonlinear dynamic system modeling problem with adaptive time delay. Then, two types of recurrent neural networks, long short-term memory neural network (LSTM NN) and gated recurrent unit neural network (GRU NN), are introduced to directly learn the dynamic prediction model from end-to-end observation data. A set of variable speed hole milling experiments are carried out to construct a cutting excitation dataset under time-varying working conditions for comparative verifications. For the X-axis with more complex dynamic characteristics, LSTM NN has better prediction performance, with a relative root mean square error of 17.62%. For the Y-axis with relatively more simple dynamic characteristics, GRU NN has better prediction performance, with a relative root mean square error of 11.74%.

Key words: cutting force, machining state monitoring, adaptive machining, recurrent neural network, servo monitoring signal

中图分类号:

TG54

程英豪, 刘长青, 庄其扬, 李光旭, 郝小忠. 基于循环神经网络的切削力无传感器监测[J]. 机械工程学报, 2025, 61(6): 14-23.

CHENG Yinghao, LIU Changqing, ZHUANG Qiyang, LI Guangxu, HAO Xiaozhong. Sensorless Cutting Force Monitoring Based on Recurrent Neural Network[J]. Journal of Mechanical Engineering, 2025, 61(6): 14-23.

导出引用管理器 EndNote|Reference Manager|ProCite|BibTeX|RefWorks

链接本文: http://www.cjmenet.com.cn/CN/10.3901/JME.2025.06.014

http://www.cjmenet.com.cn/CN/Y2025/V61/I6/14

参考文献

[1] TETI R，JEMIELNIAK K，O’DONNELL G，et al. Advanced monitoring of machining operations[J]. CIRP Annals，2010，59(2)：717-739.
[2] 滕洪钊，邓朝晖，吕黎曙，等. 多传感器信息融合的加工过程状态监测研究[J]. 机械工程学报，2022，58(6)：26-41. TENG Hongzhao，DENG Zhaohui，LÜ Lishu，et al. Research of process condition monitoring based on multi-sensor information fusion[J]. Journal of Mechanical Engineering，2022，58(6)：26-41.
[3] LI X B，LIU X L，YUE C X，et al. Systematic review on tool breakage monitoring techniques in machining operations[J]. International Journal of Machine Tools and Manufacture，2022，176：103882.
[4] 王志学，刘献礼，李茂月，等. 切削加工颤振智能监控技术[J]. 机械工程学报，2020，56(24)：1-23. WANG Zhixue，LIU Xianli，LI Maoyue，et al. Intelligent monitoring and control technology of cutting chatter[J]. Journal of Mechanical Engineering，2020，56(24)：1-23.
[5] ZHANG X，GAO Y，GUO Z C，et al. Physical model-based tool wear and breakage monitoring in milling process[J]. Mechanical Systems and Signal Processing，2023，184：109641.
[6] 李益文，邓朝晖，刘涛，等. 切削加工过程中颤振在线监测研究综述[J]. 航空学报，2023，44(11)：027562. LI Yiwen，DENG Zhaohui，LIU Tao，et al. Review on on-line monitoring of chatter in cutting process[J]. Acta Aeronautica et Astronautica Sinica，2023，44(11)：027562.
[7] 吴宝海，张阳，郑志阳，等. 数控加工进给速度参数优化研究现状与展望[J]. 航空学报，2022，43(4)：525467. WU Baohai，ZHANG Yang，ZHENG Zhiyang，et al. Review and prospects of feedrate optimization in CNC machining[J]. Acta Aeronautica et Astronautica Sinica，2022，43(4)：525615.
[8] 张阳，吴宝海，夏卫红，等. 变切深工况下恒定切削力约束的多目标进给量优化方法[J]. 机械工程学报，2021，57(5)：242-250. ZHANG Yang，WU Baohai，XIA Weihong，et al. Multi-objective feed optimization with constant cutting force constraints under variable cutting depth[J]. Journal of Mechanical Engineering，2021，57(5)：242-250.
[9] 潘泽民. CFRP/Ti复合结构螺旋铣孔自动控制技术研究[D]. 杭州：浙江大学，2016. PAN Zemin. Study on automatic control technology of helical milling on CFRP/Ti composite structures[D]. Hangzhou：Zhejiang University，2016.
[10] ALTINTAS Y. Prediction of cutting forces and tool breakage in milling from feed drive current measurements[J]. Journal of Engineering for Industry，1992，114(4)：386-392.
[11] KIM G D，CHU C N. Indirect cutting force measurement considering frictional behaviour in a machining centre using feed motor current[J]. The International Journal of Advanced Manufacturing Technology，1999，15：478-484.
[12] KIM T，WOO J，SHIN D，et al. Indirect cutting force measurement in multi-axis simultaneous NC milling processes[J]. International Journal of Machine Tools and Manufacture，1999，39：1717-1731.
[13] ASLAN D，ALTINTAS Y. Prediction of cutting forces in five-axis milling using feed drive current measurements[J]. IEEE/ASME Transactions on Mechatronics，2018，23(2)：833-844.
[14] CHENG Y H，LI Y G，LIU X，et al. Mechanism-based structured deep neural network for cutting force forecasting using CNC inherent monitoring signals[J]. IEEE/ASME Transactions on Mechatronics，2022，27(4)：2235-2245.
[15] YAMADA Y，KAKINUMA Y. Sensorless cutting force estimation for full-closed controlled ball-screw-driven stage[J]. The International Journal of Advanced Manufacturing Technology，2016，87：3337-3348.
[16] YAMADA Y，YAMATO S，KAKINUMA Y. Mode decoupled and sensorless cutting force monitoring based on multi-encoder[J]. The International Journal of Advanced Manufacturing Technology，2017，92：4081-4093.
[17] YAMATO S，KAKINUMA Y. Precompensation of machine dynamics for cutting force estimation based on disturbance observer[J]. CIRP Annals，2020，69(1)：333-336.
[18] WANG L H. From intelligence science to intelligent manufacturing[J]. Engineering，2019，5：615-618.
[19] HEYDARZADEH M S，REZAEI S M，AZIZI N，et al. Compensation of friction and force ripples in the estimation of cutting forces by neural networks[J]. Measurement，2018，114：354-364.
[20] PENG D F，LI H K，DAI Y B，et al. Prediction of milling force based on spindle current signal by neural networks[J]. Measurement，2022，205：112153.
[21] 刘丽兰，刘宏昭，吴子英，等. 机械系统中摩擦模型的研究进展[J]. 力学进展，2008，38(2)：201-213. LIU Lilan，LIU Hongzhao，WU Ziying，et al. An overview of friction models in mechanical systems[J]. Advances in Mechanics，2008，38(2)：201-213.
[22] 杨特，杨智春，梁舒雅，等. 平稳随机载荷的信号特征提取与深度神经网络识别[J]. 航空学报，2022，43(9)：225952. YANG Te，YANG Zhichun，LIANG Shuya，et al. Feature extraction and identification of stationary random dynamic load using deep neural network[J]. Acta Aeronautica et Astronautica Sinica，2022，43(9)：225952.
[23] GERS F A，SCHMIDHUBER J，CUMMINS F. Learning to forget：Continual prediction with LSTM[J]. Neural computation，2000，12(10)：2451-2471.
[24] HU C X，OU T S，CHANG H N，et al. Deep GRU neural network prediction and feedforward compensation for precision multiaxis motion control systems[J]. IEEE/ASME Transactions on Mechatronics，2020，25(3)：1377-1388.