Research on precision inference of compound biaxial turntable based on Bayesian network
-
摘要: 为精准预测复合双轴转台精度,分析了多因素对系统精度影响规律。利用复合双轴转台统计系统实验数据,学习构建以内轴精度、外轴精度、自准直精度和待检编码器精度为主要节点的贝叶斯网络结构;在Netica中建立系统精度推理模型,并通过证据敏感性分析和平均绝对误差(MAE)分析验证贝叶斯网络(BN)模型的有效性;运用自学习贝叶斯网络的概率推理,分析主要目标节点各变量的后验概率变化,对系统精度变化规律进行原因诊断和支持解释。研究结果表明:复合双轴转台精度自学习BN模型能够实现系统精度准确推理预测,系统精度超差的MAE值基本稳定在5%以内,且角度间隔0.125°和时间间隔20 s为系统最优控制参数,为贝叶斯网络技术在复合双轴转台精度推理中的应用提供了参考。Abstract: In order to accurately predict the accuracy of the compound biaxial turntable, the influencing rule of multiple factors on the system accuracy was analyzed. By using the statistical experimental data of the compound biaxial turntable, the system learns and constructs the Bayesian network structure with the inner axis accuracy, outer axis accuracy, auto-collimation accuracy and the accuracy of the encoder to be tested as the main nodes. The inference model of the system detection accuracy was established in Netica software, and the validity of the Bayesian network (BN) model was verified through evidence sensitivity analysis and mean absolute error (MAE) analysis. The self-learning Bayesian network probabilistic inference is used to analyze the posterior probability changes of each variable of the main target node, and to diagnose and support the reason for the change of the system accuracy. The research results show that the self-learning BN model with compound biaxial turntable accuracy can achieve accurate inference and prediction with system accuracy. And the MAE value of the system accuracy out of tolerance is basically stable within 5%, and the angular interval of 0.125° and the time interval of 20 s are the optimal control parameters of the system, which provides a reference for the application of Bayesian network technology in the accuracy inference of the compound biaxial turntable.
-
Key words:
- compound biaxial turntable /
- Bayesian network /
- Netica /
- precision inference
-
表 1 “系统精度”父节点敏感性分析结果
节点 互信息值 百分比/(%) 方差值 角度间隔(AI) 0.368 65 36.87 0.102 83 时间间隔(TI) 0.137 12 13.71 0.043 70 内轴精度(IA) 0.022 81 2.28 0.007 23 外轴精度(OA) 0.020 68 2.07 0.006 42 自准直精度(AA) 0.054 80 5.48 0.015 16 编码器精度(EA) 0.057 15 5.72 0.017 05 表 2 Netica后验概率与实验统计概率
% 变量 类型 实验统计数据 Netica预测数据 y1 n1 n2 n3 n4 y1 n1 n2 n3 n4 角度间隔(AI) (A1I) 64.22 36.23 7.81 4.12 6.54 65.6 23.5 4.35 2.17 11.1 (A2I) 22.14 18.35 10.62 2.57 20.15 23.4 11.8 4.35 2.17 11.1 (A3I) 4.51 24.51 16.15 4.96 17.23 7.81 52.9 17.4 2.17 11.1 (A4I) 7.90 10.17 48.17 4.86 6.58 1.56 5.88 69.6 2.17 11.1 (A5I) 1.23 10.74 17.25 83.49 49.50 1.56 5.88 4.35 91.3 55.6 时间间隔(TI) (T1I) 37.36 35.73 38.64 35.47 36.92 40.0 38.7 53.3 31.7 40.0 (T2I) 28.40 32.16 32.42 31.05 36.81 29.0 30.6 23.3 33.9 32.0 (T3I) 34.24 32.11 28.94 33.48 26.67 31.0 30.6 23.5 34.4 28.0 表 3 自学习BN模型预测概率与实验概率MAE对比
y1 n1 n2 n3 n4 角度间隔/(°) 0.025 2 0.113 6 0.090 6 0.031 2 0.060 7 时间间隔/s 0.047 2 0.040 6 0.194 5 0.050 2 0.071 7 -
[1] 董凯炎, 史妹倩, 刘杰, 等. 一种光电编码器精度自动检测系统[J]. 传感技术学报, 2019, 32(3): 463-468. doi: 10.3969/j.issn.1004-1699.2019.03.024 [2] 李家琨, 冯其波, 包传辰, 等. 转轴转角定位误差测量方法与误差分析[J]. 红外与激光工程, 2019, 48(2): 205-210. [3] LI K, Ding H C, Cao G H, et al. Research on high precision error measurement system of angular displacements based on reciprocal roll angles[J]. Optical Engineering, 2020, 59(12): 124110. [4] 王福全, 王环, 谢志江, 等. 精密转台角分度误差补偿[J]. 光学精密工程, 2017, 25(8): 2165-2172. [5] Dong J, Wan Q H, Yu H, et al. Automatic detection system of fault code for small size absolute photoelectric encoder[J]. Chinese Optics, 2016, 9(6): 695-703. [6] 王文君, 段晓君, 朱炬波, 等. 某型弹载雷达测角系统误差模型辨识方法[J]. 兵工学报, 2014, 35(2): 273-279. doi: 10.3969/j.issn.1000-1093.2014.02.021 [7] 邬昌军, 郑明明, 王巧花, 等. 横梁移动式五轴龙门铣床摆动轴与位置相关的几何误差检测与辨识[J]. 制造技术与机床, 2022(9): 173-180. [8] 项四通. 五轴数控机床空间误差测量、建模与补偿技术研究[D]. 上海: 上海交通大学, 2016. [9] 向华, 余金舫, 王超, 等. 基于神经网络的机床误差数据拟合与预测方法[J]. 组合机床与自动化加工技术, 2019(12): 44-47,51. doi: 10.13462/j.cnki.mmtamt.2019.12.011 [10] 熊青春, 王家序, 周青华, 等. 融合机床精度与工艺参数的铣削误差预测模型[J]. 航空学报, 2018, 39(8): 272-280. doi: 10.7527/S1000-6893.2018.21713 [11] 辛宗霈, 冯显英, 杜付鑫, 等. 基于BP神经网络的机床热误差建模与分析[J]. 组合机床与自动化加工技术, 2019(8): 39-43. doi: 10.13462/j.cnki.mmtamt.2019.08.012 [12] 王艺玮, 邓蕾, 郑联语, 等. 基于多通道融合及贝叶斯理论的刀具剩余寿命预测方法[J]. 机械工程学报, 2021, 57(13): 214-224. [13] Li K, Cao G H, Ding H C, et al. Research on the new type of error detection platform for photoelectric encoder[C]. 2021 International Conference on Mechanical Engineering and Advanced Control (MEAC2021), 2021, (2029): 012140.