Application of improved gray wolf algorithm to optimize gray forecasting model in CNC machine tools
-
摘要: 针对传统灰色预测模型因背景值选取带来的预测误差较大的问题,提出一种与改进灰狼算法相结合的故障预测模型。设计一种改进的灰狼算法,对基本灰狼算法的算法参数通过非线性策略进行改善,并用于优化灰色预测模型中的背景值,从而获得最优预测模型。以某一型号数控车床主轴的8个故障数据为例,将该预测模型与其他灰色改进预测模型进行对比验证,结果显示此模型与原始数据拟合度及稳定性最好。Abstract: Aiming at the problem of large prediction errors caused by the selection of background values in the traditional gray prediction model, a fault prediction model combined with the improved gray wolf algorithm is proposed. An improved gray wolf algorithm is designed to improve the algorithm parameters of the basic gray wolf algorithm and used to optimize the background value in the gray prediction model to obtain the optimal prediction model. Taking 8 fault data of CNC lathe spindle as an example, the prediction accuracy of the proposed method is compared with other gray models to verify the fitness and stability of the method.
-
Key words:
- CNC machine tool /
- gray theory /
- improved gray wolf algorithm /
- fault prediction
-
表 1 不同模型预测值及相对误差
序号 实际数据/h 改进的GGWO模型 GGWO模型 GPSO模型 预测值/h 误差/(%) 预测值/h 误差/(%) 预测值/h 误差/(%) 1 118.99 119.0 - 119.0 - 119.0 - 2 200.39 205.56 2.58 207.2 3.40 206.0 2.90 3 293.13 297.6 1.51 293.6 0.15 304.7 3.95 4 400.89 392.5 2.10 386.2 3.67 403.8 0.72 5 539.50 517.7 2.24 508.0 4.05 535.0 1.05 6 688.98 682.8 0.90 668.3 3.00 709.0 2.90 7 899.04 900.5 0.16 879.2 2.21 939.4 4.49 8 1 187.45 1 187.7 0.002 4 1 156.6 2.26 1 244.8 4.83 表 2 各模型评价指标计算结果
模型名称 改进GGWO模型 GGWO模型 GPSO模型 RMSE 8.89 19.45 26.25 MAE 5.97 15.61 17.79 -
[1] 刘强. 数控机床发展历程及未来趋势[J]. 中国机械工程, 2021, 32(7): 757-770. doi: 10.3969/j.issn.1004-132X.2021.07.001 [2] 代愽超, 张英杰, 李阳帆, 等. 基于性能退化预测的数控机床预防维修方法[J]. 中国机械工程, 2019, 30(17): 2122-2128. doi: 10.3969/j.issn.1004-132X.2019.17.015 [3] 王鑫, 吴际, 刘超, 等. 基于LSTM循环神经网络的故障时间序列预测[J]. 北京航空航天大学学报, 2018, 44(4): 772-784. [4] 李雪, 王鹤, 邵秋萍, 等. 基于遗传神经网络的数控机床故障预测[J]. 现代制造技术与装备, 2016(10): 17-18. doi: 10.3969/j.issn.1673-5587.2016.10.008 [5] 王强军. 基于指数平滑方法的丝杠轴承故障时间预测[J]. 制造技术与机床, 2014(4): 83-86. doi: 10.3969/j.issn.1005-2402.2014.04.027 [6] Zhou P, Ang B W, Poh K L. A trigonometric grey prediction approach to forecasting electricity demand[J]. Energy, 2006, 31(14): 2839-2847. doi: 10.1016/j.energy.2005.12.002 [7] 徐文, 薛玉霞, 申桂香. 基于粒子群优化改进灰色模型的数控机床故障预测[J]. 制造业自动化, 2012, 34(8): 52-55. doi: 10.3969/j.issn.1009-0134.2012.4(x).15 [8] 丁海峰, 高凯, 罗娟, 等. 基于GM(1, 1)灰色预测模型的我国民营医院发展趋势预测[J]. 医学与社会, 2021, 34(3): 1-6. [9] 丁松. 灰色预测模型优化及其应用研究[D]. 南京: 南京航空航天大学, 2018. [10] Yao A, Chi S C, Chen J H. An improved grey-based approach for electricity demand forecasting[J]. El-ectric Power Systems Research, 2003, 67(3): 217-224. doi: 10.1016/S0378-7796(03)00112-3 [11] Mirjalili S, Mirjalili S M, Lewis A. Grey wolf optimization[J]. Advances in Engineering Software, 2014, 69(7): 46-61. [12] Chiu C Y, Shih P C, Li X C. A dynamic adjusting novel global harmo ny search for continuous optimization problems[J]. Symmetry, 2018, 10(8): 337. doi: 10.3390/sym10080337