Life Prediction of Milling Cutters Combining AMPSO with SVR
-
摘要: 针对支持向量回归机在预测铣刀寿命时惩罚参数 和核函数参数 难确定、不同的参数设置对预测效果影响较大的问题,提出了自适应变异粒子群算法。在支持向量回归算法的基础上,引入AMPSO优化SVR参数,建立AMPSO与SVR相结合的数控铣刀寿命预测模型。通过硬质合金钢铣刀铣削的实验验证表明,相比于网格搜索法和神经网络算法,AMPSO-SVR算法在测试样本集的平均相对预测误差低至0.72%,相较前两者预测误差更小,可准确预测数控铣刀寿命,为数控加工过程中的换刀决策提供依据。
-
关键词:
- 铣刀 /
- 支持向量回归机 /
- 自适应变异粒子群算法 /
- 寿命预测
Abstract: Aiming at the problems that it is difficult to determine the penalty parameter and kernel function parameter , and different parameters have a great influence on the prediction effect with the support vector regression to predict the milling cutter’s life, the adaptive mutation particle swarm optimizationalgorithm was proposed. On the basis of the support vector regression algorithm, AMPSO (Adaptive Mutation Particle Swarm Optimization) was used to optimize the SVR (Support Vector Regression) parameters, and a life prediction model for milling cutter combining AMPSO with SVR was established. The results show that the average relative prediction error of the AMPSO-SVR algorithm in the test sample set is as low as 0.72%, which is smaller than that of the grid search method and the neural network algorithm. It can accurately predict the milling cutter’s life, which can provide a basis for tool change decisions during the machining process. -
表 1 实验样本数据(训练 + 测试)
样本 铣削速度/(r·min−1 ) 铣削深度/mm 铣削宽度/mm 铣刀直径/mm 铣刀齿数/个 每齿进给量/(mm·z−1) 实际寿命/h 1 197 1 40 60 3 0.08 80 2 182 2 60 80 4 0.10 90 3 164 2 80 100 5 0.12 105 4 124 5 120 160 8 0.16 155 5 115 6 120 180 9 0.18 170 6 90 8 200 250 12 0.18 245 7 78 12 220 300 13 0.18 290 8 106 8 140 200 10 0.15 200 9 150 4 80 120 6 0.14 120 10 132 4 100 140 7 0.15 135 表 2 测试样本集上的铣刀寿命预测结果
样本 实际寿命/h 预测寿命/h 绝对误差/h 相对误差/% 8 200 198.8257 1.1743 0.5872 9 120 119.1992 0.8008 0.6673 10 135 136.2269 1.2269 0.9088 表 3 归一化方式对比
样本 不归一化 [−1,1]归一化 [0,1]归一化 8 19.9984% 0.5872% 2.0883% 9 45.8361% 0.6673% 2.0131% 10 18.5210% 0.9088% 4.1526% 表 4 算法预测结果对比
样
本实际
寿命/hAMPSO-SVR GS-SVR BP神经网络 预测
寿命/h相对
误差/%训练
时间/s预测
寿命/h相对
误差/%训练
时间/s预测
寿命/h相对
误差/%训练
时间/s8 200 198.8257 0.5872 0.87 202.0369 1.0185 0.43 188.6100 5.6950 2.10 9 120 119.1992 0.6673 0.87 120.7146 0.5955 0.43 112.2208 6.4827 2.10 10 135 136.2269 0.9088 0.87 137.8866 2.1382 0.43 124.8938 7.4861 2.10 -
[1] KONG D D, CHEN Y J, LI N, et al. Tool wear monitoring based on kernel principal component analysis and v-support vector regression[J]. The International Journal of Advanced Manufacturing Technology, 2017, 89(1-4): 175-190. doi: 10.1007/s00170-016-9070-x [2] 潘柏松, 俞铭杰, 项涌涌, 等. 考虑刀具磨损的铣削加工精度可靠性分析及工艺优化设计[J]. 计算机集成制造系统, 2020, 26(11): 2982-2991. doi: 10.13196/j.cims.2020.11.008PAN B S, YU M J, XIANG Y Y, et al. Accuracy reliability analysis and process optimization design of milling processing considering tool wear[J]. Computer Integrated Manufacturing Systems, 2020, 26(11): 2982-2991. (in Chinese) doi: 10.13196/j.cims.2020.11.008 [3] Taylor F W. On the art of cutting metals[M]. American Society of Mechanical Engineers, 1906. [4] Hastings W F, Mathew P, Oxley L B. Estimated cutting temperatures their use as a predictor of tool performance when machining plain carbon steels[C]. Proceedings of thr 20th International MTDR Conference, 1979 [5] 辛红敏, 吴华伟. 整体叶盘盘铣开槽加工刀具寿命预测[J]. 河南科技大学学报(自然科学版), 2020, 41(2): 16-20.XIN H M, WU H W. Tool life prediction in disc milling grooving of blisk[J]. Journal of Henan University of Science and Technology (Natural Science), 2020, 41(2): 16-20. (in Chinese) [6] WU D Z, JENNINGS C, TERPENNY J, et al. A comparative study on machine learning algorithms for smart manufacturing: tool wear prediction using random forests[J]. Journal of Manufacturing Science and Engineering, 2017, 139(7): 071018. doi: 10.1115/1.4036350 [7] 曾晓雪, 吉卫喜, 徐杰. 基于CPSO-BP的刀具寿命预测算法[J]. 组合机床与自动化加工技术, 2020(8): 57-59.ZENG X X, JI W X, XU J. Research on tool life prediction method based on CPSO-BP[J]. Modular Machine Tool & Automatic Manufacturing Technique, 2020(8): 57-59. (in Chinese) [8] 李鑫, 史振宇, 蒋森河, 等. 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.LI X, SHI Z Y, JIANG S H, et al. Artificial neural network predicts tool wear and cutting force[J]. Control Theory & Applications, 2018, 35(12): 1731-1737. (in Chinese) [9] CHENG M H, JIAO L, SHI X C, et al. An intelligent prediction model of the tool wear based on machine learning in turning high strength steel[J]. Proceedings of the Institution of Mechanical Engineers, Part B:Journal of Engineering Manufacture, 2020, 234(13): 1580-1597. doi: 10.1177/0954405420935787 [10] 程灿, 李建勇, 徐文胜, 等. 基于支持向量机与粒子滤波的刀具磨损状态识别[J]. 振动与冲击, 2018, 37(17): 48-55.CHENG C, LI J Y, XU W S, et al. Tools' wear state recognition based on support vector machine and particle filtering[J]. Journal of Vibration and Shock, 2018, 37(17): 48-55. (in Chinese) [11] BENKEDJOUH T, MEDJAHER K, ZERHOUNI N, et al. Health assessment and life prediction of cutting tools based on support vector regression[J]. Journal of Intelligent Manufacturing, 2015, 26(2): 213-223. doi: 10.1007/s10845-013-0774-6 [12] 江鸿怀, 金晓怡, 邢亚飞, 等. 基于粒子群优化算法的五自由度机械臂轨迹规划[J]. 机械设计与研究, 2020, 36(1): 107-110. doi: 10.13952/j.cnki.jofmdr.2020.0022JIANG H H, JIN X Y, XING Y F, et al. Five-degree-of-freedom manipulator trajectory planning based on PSO particle algorithm[J]. Machine Design & Research, 2020, 36(1): 107-110. (in Chinese) doi: 10.13952/j.cnki.jofmdr.2020.0022 [13] 杨旭, 邱明, 陈立海, 等. 基于PSO-RWE的自适应小波阈值函数滚动轴承振动信号去噪方法[J]. 航空动力学报, 2020, 35(11): 2339-2347.YANG X, QIU M, CHEN L H, et al. Adaptive wavelet threshold function based on PSO-RWE for vibration signal denoising of rolling bearing[J]. Journal of Aerospace Power, 2020, 35(11): 2339-2347. (in Chinese) [14] DENG W, YAO R, ZHAO H M, et al. A novel intelligent diagnosis method using optimal LS-SVM with improved PSO algorithm[J]. Soft Computing, 2019, 23(7): 2445-2462. doi: 10.1007/s00500-017-2940-9 [15] 薛宏. 企业级刀具全生命周期管理系统研发[D]. 重庆: 重庆大学, 2015XUE H. The research on enterprise tool lifecycle management system[D]. Chongqing: Chongqing University, 2015. (in Chinese) [16] 周志华. 机器学习[M]. 北京: 清华大学出版社, 2016ZHOU Z H. Machine learning[M]. Beijing: Tsinghua University Press, 2016. (in Chinese) [17] 黄媛, 孙树栋, 李兢尧. 基于ACO-BP神经网络的刀具寿命预测[J]. 机械科学与技术, 2009, 28(11): 1517-1521. doi: 10.3321/j.issn:1003-8728.2009.11.027HUANG Y, SUN S D, LI J Y. Prediction of cutting tool life based on ACO-BP neural network[J]. Mechanical Science and Technology for Aerospace Engineering, 2009, 28(11): 1517-1521. (in Chinese) doi: 10.3321/j.issn:1003-8728.2009.11.027