Study on Tool Condition-Integrated Online Optimization of Process Parameters
-
摘要:
随着现代制造业对加工质量与生产效率要求的不断提升，刀具磨损已成为影响表面粗糙度的关键制约因素. 传统的刀具状态监测及工艺参数优化方法多基于经验模型或静态优化策略，难以适应多变量、动态变化的复杂加工环境. 针对这一挑战，创新性地提出了一种融合多尺度分布比（MSDR）与贝叶斯多臂老虎机（BMAB）的工艺参数在线优化方法，将刀具状态纳入工艺参数优化框架中；结合贝叶斯优化和多臂老虎机策略，在动态加工环境中实现了工艺参数的实时调整，通过保证加工效率最大化的同时，维持加工过程的稳定性和精确性. 研究结果表明：与主流方法相比，MSDR在刀具状态监测中展现出优异的精度和稳定性，其MAE、SMER和RMSE分别达到0.145、0.258和0.194；BMAB在切削效率优化和计算时效性方面亦表现出色，分别达到
2305 mm3/min和2.92 s，显著优于传统方法. 因此，考虑刀具状态的工艺参数在线优化技术为高精度制造提供了一条全新的技术路线.Abstract:As demands for manufacturing quality and production efficiency continue to rise in modern industry, tool wear has emerged as a critical constraint affecting surface roughness. Traditional tool condition monitoring and process parameter optimization methods are often based on empirical models or static optimization strategies, limiting their adaptability to complex, dynamically changing, multivariable environments. In response, this study proposes an innovative approach integrating multi-scale distribution ratio (MSDR) with Bayesian multi-armed bandit (BMAB) for process parameter online optimization, incorporating real-time tool condition data into the optimization framework. Additionally, by combining Bayesian optimization and multi-armed bandit strategies, this method enables real-time adjustments to process parameters in dynamic manufacturing environments, effectively balancing exploration and exploitation to maximize machining efficiency. Compared to mainstream methods, MSDR demonstrates exceptional precision and stability in tool condition monitoring, achieving MAE, SMER, and RMSE values of 0.145, 0.258, and 0.194, respectively. BMAB also performs exceptionally in optimizing cutting efficiency and computational effectiveness, achieving
2305 mm3/min and a runtime of 2.92 seconds, respectively. Therefore, tool state-aware online optimization of process parameters presents a novel and promising technical pathway for high-precision manufacturing. -
图 1 考虑刀具状态的工艺参数在线优化流程
Figure 1. Flow chart of tool condition-integrated online optimization of process parameters.
图 10 工艺参数优化模型的时效性评估
Figure 10. The timeliness evaluation of each process parameter optimization model.
表 1 铣刀磨损数据集详细信息
Table 1. Detailed information of milling cutter wear data set.
参数 取值 铣刀型号 APMT1135 铣刀类型 硬质合金刀 被试材料 45钢 机床型号 VMC850 机床功率/KW 7.5 主轴转速范围/（r•min−1） 20～ 8000 切削进给速度/（mm•min−1） 1～ 10000 环境温度要求/（℃） 0～4 相对湿度/% 20～80 
下载: 导出CSV
表 2 铣刀磨损数据集详细信息
Table 2. Detailed information of milling cutter wear data set.
实验次数/
次切深/
mm转速/
（r•s−1）进给/
（mm•s−1）后刀面磨
损/μm表面粗糙
度/μm1.0 1.0 66.7 1.11 86.0 2.138 2.0 1.0 58.3 0.97 224.4 2.920 3 1.0 33.3 0.56 171.0 1.577 4 1.0 33.3 0.56 48.5 1.897 5 1.0 58.3 0.97 210.2 2.599 6 1.0 58.3 0.97 67.0 0.945 7 1.0 33.3 0.56 171.0 3.714 8 1.0 83.3 1.39 285.0 0.486 9 1.0 41.7 0.69 160.3 1.728 10 1.0 83.3 1.39 162.5 1.697 11 1.0 66.7 1.11 85.6 2.561 12 1.5 83.3 1.39 104.8 1.315 13 1.5 50.0 0.83 118.3 2.996 14 1.5 66.7 1.11 45.7 0.459 15 1.5 66.7 1.11 123.0 1.323 16 1.5 75.0 1.25 168.6 1.250 17 1.5 66.7 1.11 48.3 0.581 18 1.5 75.0 1.25 56.4 1.106 19 1.5 50.0 0.83 69.8 0.829 20 1.5 50.0 0.83 214.9 1.241 21 1.5 33.3 0.56 56.4 3.804 22 1.5 50.0 0.83 212.3 1.353 23 1.5 41.7 0.69 112.5 2.429 24 1.5 33.3 0.56 106.8 2.579 25 2.0 58.3 0.97 131.8 1.376 26 2.0 66.7 1.11 220.4 1.244 27 2.0 41.7 0.69 118.1 3.020 28 2.0 50.0 0.83 94.0 2.966 29 2.0 58.3 0.97 126.0 1.229 30 2.0 50.0 0.83 270.5 1.595
下载: 导出CSV
表 3 各健康指标性能评价
Table 3. Performance evaluation of His.
模型 评价指标 MAE SMER RMSE ED-HI 0.216 0.358 0.269 GMM-HI 0.424 0.360 0.512 MMD-HI 0.489 0.574 0.53 RMS-HI 0.196 0.340 0.246 RS-HI 0.183 0.338 0.232 MSDR-HI 0.145 0.258 0.194
下载: 导出CSV
-
[1] 刘洪成, 袁德志, 朱锟鹏. 基于高斯过程潜力模型的刀具磨损预测[J]. 机械工程学报, 2023, 59(17): 310-324. doi: 10.3901/JME.2023.17.310LIU Hongcheng, YUAN Dezhi, ZHU Kunpeng. Tool wear prediction based on Gaussian process latent force model[J]. Journal of Mechanical Engineering, 2023, 59(17): 310-324. doi: 10.3901/JME.2023.17.310 [2] YOU Z C, GAO H L, GUO L, et al. On-line milling cutter wear monitoring in a wide field-of-view camera[J]. Wear, 2020, 460: 203479. [3] 高宏力, 孙弋, 郭亮, 等. 机械加工质量预测研究现状与发展趋势[J]. 西南交通大学学报, 2024, 59(1): 121-141.GAO Hongli, SUN Yi, GUO Liang, et al. Research status and development trend of machining quality prediction[J]. Journal of Southwest Jiaotong University, 2024, 59(1): 121-141. [4] LE V T, HOANG L, GHAZALI M F, et al. Optimization and comparison of machining characteristics of SKD61 steel in powder-mixed EDM process by TOPSIS and desirability approach[J]. The International Journal of Advanced Manufacturing Technology, 2024, 130(1): 403-424. [5] SUN Y, SONG H L, GAO H L, et al. Interpretable tool wear monitoring: Architecture with large-scale CNN and adaptive EMD[J]. Journal of Manufacturing Systems, 2025, 78: 294-307. doi: 10.1016/j.jmsy.2024.12.001 [6] BERUVIDES G, CASTAÑO F, QUIZA R, et al. Surface roughness modeling and optimization of tungsten–copper alloys in micro-milling processes[J]. Measurement, 2016, 86: 246-252. doi: 10.1016/j.measurement.2016.03.002 [7] YUSUP N, SARKHEYLI A, ZAIN A M, et al. Estimation of optimal machining control parameters using artificial bee colony[J]. Journal of Intelligent Manufacturing, 2014, 25(6): 1463-1472. doi: 10.1007/s10845-013-0753-y [8] MOREIRA L C, LI W D, LU X, et al. Supervision controller for real-time surface quality assurance in CNC machining using artificial intelligence[J]. Computers & Industrial Engineering, 2019, 127: 158-168. [9] 李聪波, 余必胜, 肖溱鸽, 等. 考虑刀具磨损的数控车削批量加工工艺参数节能优化方法[J]. 机械工程学报, 2021, 57(1): 217-229. doi: 10.3901/JME.2021.01.217LI Congbo, YU Bisheng, XIAO Qinge, et al. A cutting parameter energy-saving optimization method for CNC turning batch processing considering tool wear[J]. Journal of Mechanical Engineering, 2021, 57(1): 217-229. doi: 10.3901/JME.2021.01.217 [10] 张维煜, 李凯, 杨鑫. 基于参数优先级划分的飞轮电机多目标优化[J]. 西南交通大学学报, 2023, 58(4): 922-932. doi: 10.3969/j.issn.0258-2724.20220845ZHANG Weiyu, LI Kai, YANG Xin. Multi-objective optimization for flywheel motors based on parameter priority division[J]. Journal of Southwest Jiaotong University, 2023, 58(4): 922-932. doi: 10.3969/j.issn.0258-2724.20220845 [11] KUNTOĞLU M, ASLAN A, SAĞLAM H, et al. Optimization and analysis of surface roughness, flank wear and 5 different sensorial data via tool condition monitoring system in turning of AISI 5140[J]. Sensors, 2020, 20(16): 4377. doi: 10.3390/s20164377 [12] NG J J, ZHONG Z W, LIU T I. Prediction of roughness heights of milled surfaces for product quality prediction and tool condition monitoring[J]. Journal of Materials and Applications, 2019, 8(2): 97-104. doi: 10.32732/jma.2019.8.2.97 [13] WANG Y H, WANG Y W, ZHENG L Y, et al. Online surface roughness prediction for assembly interfaces of vertical tail integrating tool wear under variable cutting parameters[J]. Sensors, 2022, 22(5): 1991. doi: 10.3390/s22051991 [14] 由智超, 高宏力, 郭亮, 等. 基于改进信息熵的直接刀具状态监测设备部署[J]. 西南交通大学学报, 2024, 59(1): 160-167. doi: 10.3969/j.issn.0258-2724.20220025YOU Zhichao, GAO Hongli, GUO Liang, et al. Equipment deployment of direct tool-condition monitoring based on improved information entropy[J]. Journal of Southwest Jiaotong University, 2024, 59(1): 160-167. doi: 10.3969/j.issn.0258-2724.20220025 [15] 汪鑫, 廖小平, 刘树胜, 等. 多传感器融合下多工况刀具磨损状态预测的深度森林方法研究[J]. 仪器仪表学报, 2023, 44(9): 263-272. doi: 10.19650/j.cnki.cjsi.J2310980WANG Xin, LIAO Xiaoping, LIU Shusheng, et al. Research on the deep forest method for tool wear state prediction under multiple working conditions with multi-sensor fusion[J]. Chinese Journal of Scientific Instrument, 2023, 44(9): 263-272. doi: 10.19650/j.cnki.cjsi.J2310980 [16] 吕彪, 刘于萌. 基于贝叶斯网络的CTCS-3级列控车载系统韧性[J]. 西南交通大学学报, 2022, 57(5): 949-959. doi: 10.3969/j.issn.0258-2724.20210102LYU Biao, LIU Yumeng. Resilience assessment based on Bayesian network for on-board subsystem of CTCS-3 train control system[J]. Journal of Southwest Jiaotong University, 2022, 57(5): 949-959. doi: 10.3969/j.issn.0258-2724.20210102 [17] 邓聪颖, 叶波, 禄盛, 等. 基于切削稳定性与表面质量约束的铣削工艺参数优化研究[J]. 仪器仪表学报, 2021, 42(11): 190-199. doi: 10.19650/j.cnki.cjsi.J2108272DENG Congying, YE Bo, LU Sheng, et al. Optimization of milling process parameters considering the constraints of cutting stability and surface quality[J]. Chinese Journal of Scientific Instrument, 2021, 42(11): 190-199. doi: 10.19650/j.cnki.cjsi.J2108272 [18] WANG D, ZHONG J J, SHEN C Q, et al. Correlation dimension and approximate entropy for machine condition monitoring: Revisited[J]. Mechanical Systems and Signal Processing, 2021, 152: 107497. doi: 10.1016/j.ymssp.2020.107497 [19] STEMP W J, CHILDS B E, VIONNET S, et al. The quantification of microwear on chipped stone tools: assessing the effectiveness of root mean square roughness (rq)[J]. Lithic Technology, 2008, 33(2): 173-189. doi: 10.1080/01977261.2008.11721067 [20] DIAZ-ROZO J, BIELZA C, LARRAÑAGA P. Machine-tool condition monitoring with Gaussian mixture models-based dynamic probabilistic clustering[J]. Engineering Applications of Artificial Intelligence, 2020, 89: 103434. doi: 10.1016/j.engappai.2019.103434 [21] WANG F B, LIU J K, SHU Q L. Milling wear of carbide tool for processing nickel-based alloy in cryogenic based on the entropy change[J]. The International Journal of Advanced Manufacturing Technology, 2017, 90(5): 1703-1713. [22] LIU M J, GONG Y D, SUN J Y, et al. The accuracy losing phenomenon in abrasive tool condition monitoring and a noval WMMC-JDA based data-driven method considered tool stochastic surface morphology[J]. Mechanical Systems and Signal Processing, 2023, 198: 110410. doi: 10.1016/j.ymssp.2023.110410 [23] 贾智最, 杜茂华, 王沛鑫. 基于极限学习机与遗传算法的刀具状态监测[J/OL]. 昆明理工大学学报(自然科学版), 2024: 1-14. (2024-04-25). https://kns.cnki.net/kcms/detail/53.1223.N.20240424.1556.002.html. [24] 杨沛东, 黄华, 尉卫卫, 等. 基于注意力机制改进的PSO-BiLSTM刀具磨损预测[J/OL]. 北京航空航天大学学报, 1-12 [2024-11-07]. https://doi.org/10.13700/j.bh.1001-5965.2023.0545. [25] 徐昌鸿, 张树生, 梁嘉宸. 基于数控工艺信息挖掘的型腔特征粗加工刀具序列优化决策方法[J]. 计算机集成制造系统, 2024, 30(3): 1060-1071. doi: 10.13196/j.cims.2023.0058XU Changhong, ZHANG Shusheng, LIANG Jiachen. NC process information mining based roughing tool sequence optimization and decision approach for pocket features[J]. Computer Integrated Manufacturing Systems, 2024, 30(3): 1060-1071. doi: 10.13196/j.cims.2023.0058 -
下载:
点击查看大图
计量
- 文章访问数: 34
- HTML全文浏览量: 25
- PDF下载量: 15
- 被引次数: 0