基于知识深度置信网络的加工粗糙度预测

doi:10.3901/JME.2019.20.094

摘要/Abstract

摘要： 深度神经网络是一种具有复杂结构和多个非线性处理单元的模型，目前也已逐步被应用在工业生产过程中。但由于神经网络不可解释，不可控制的"黑箱"问题，以及海量的数据需求问题，使得深度学习在工业领域的应用仍有巨大的障碍。提出一种新的深度神经网络模型：知识深度置信网络（Knowledge-based deep belief network，KBDBN）。这种逻辑符号语言与深度神经网络的结合，不仅使得模型具有良好的模式识别性能，还可自适应地确定网络模型并具有可解释和可视化特性。进一步提出基于KBDBN的工件表面粗糙度加工过程的预测模型，实现了精确预测且有效地提取了制造过程的关键知识。试验结果证明：相较于传统机器学习器，KBDBN的网络性能更加优越，具有可解释性，可应用性更强。创新性的将符号规则与深度学习相结合并建立加工粗糙度预测模型，可以在精准预测的前提下提取工艺知识，指导加工工艺优化。

关键词: 表面粗糙度, 深度学习, 深度置信网络, 知识发现, 模式识别

Abstract: Deep neural network (DNN) is a complex structure and multiple nonlinear processing unit models, and has been gradually applied in industrial production processes. Due to the unexplained of "black box" problem, and the huge data demand problem, however, there are still huge obstacles to the application of DNNs in industrial fields. A new DNN model, knowledge-based deep belief network (KBDBN), is proposed. The combination of this logical symbol language and deep neural network not only makes the model have good pattern recognition performance, but also adaptively determines the network model and has interpretable and visual characteristics. Furthermore, the prediction model of workpiece surface roughness processing based on KBDBN is proposed, which realizes accurate prediction and effectively extracts the key knowledge of the manufacturing process. Experimental results show that compared with the traditional machine learner, KBDBN has better network performance, model interpretability and more applicability. Combines symbolic rules with deep learning and establishes a processing roughness prediction model, which can extract process knowledge and guide process optimization under the premise of accurate prediction.

Key words: surface roughness, deep learning, deep belief network, knowledge discovery, pattern recognition

中图分类号:

TG84

刘国梁, 余建波. 基于知识深度置信网络的加工粗糙度预测[J]. 机械工程学报, 2019, 55(20): 94-106.

LIU Guoliang, YU Jianbo. Machining Roughness Prediction Based on Knowledge-based Deep Belief Network[J]. Journal of Mechanical Engineering, 2019, 55(20): 94-106.

参考文献

[1] TSAI Y H,CHEN J C,LOU S J. An in-process surface recognition system based on neural networks in end milling cutting operations[J]. International Journal of Machine Tools & Manufacture,1999,39(4):583-605.
[2] LELA B,BAJIC D,JOZIC S. Regression analysis,support vector machines,and Bayesian neural network approaches to modeling surface roughness in face milling[J]. International Journal of Advanced Manufacturing Technology,2009,42(11-12):1082-1088.
[3] 张永宏,胡德金,张凯,等. 基于进化神经网络的曲面磨削表面粗糙度预测[J]. 上海交通大学学报,2005,39(3):373-376. ZHANG Yonghong,HU Dejin,ZHANG Kai,et al. Prediction of the surface roughness in curve grinding based on evolutionary neural networks[J]. Journal of Shanghai Jiao Tong University,2005,39(3):373-376.
[4] 李晓梅,丁宁,朱喜林. 表面粗糙度模糊神经网络在线辨识模型[J]. 机械工程学报,2007,43(3):212-217. LI Xiaomei,DING Ning,ZHU Xilin. Surface roughness fuzzy neural network identification model wood online[J]. Journal of Mechanical Engineering,2007,43(3):212-217.
[5] WEN Long,LI Xinyu,GAO Liang,et al. Surface roughness prediction in end milling by using predicted point oriented local linear estimation method[J]. The International Journal of Advanced Manufacturing Technology,2016,84(9-12):2523-2535.
[6] 侯文擎,叶鸣,李巍华. 基于改进堆叠降噪自编码的滚动轴承故障分类[J]. 机械工程学报,2018,54(7):87-96. HOU Wenqin,YE Ming,LI Weihua. Rolling element bearing fault classification using improved stacked de-noising auto-encoders[J]. Journal of Mechanical Engineering,2018,54(7):87-96.
[7] JING Luyang,ZHAO Ming,LI Pin,et al. A convolutional neural network based feature learning and fault diagnosis method for the condition monitoring of gearbox[J]. Measurement,2017,2017(11):1-10.
[8] 孙文珺,邵思羽,严如强. 基于稀疏自动编码深度神经网络的感应电机故障诊断研究[J]. 机械工程学报,2016,52(9):65-71. SUN Wenjun,SHAO Siyu,YAN Ruqiang. Induction motor fault diagnosis based on sparse auto-encoder deep neural network[J]. Journal of Mechanical Engineering,2016,52(9):65-71.
[9] 李梦园. 深度学习算法在表面缺陷识别中的应用研究[D]. 杭州:浙江工业大学,2015. LI Mengyuan. Research and application of deep learning algorithm in surface defect recognition[D]. Hangzhou:Zhejiang University of Technology,2015.
[10] HINTON G E,OSINDERO S,TEH Y W. A fast learning algorithm for deep belief nets[J]. Neural Computation.,2006,18(7):1527-1554.
[11] SMOLENSK P. Information processing in dynamical systems:Foundations of harmony theory[C]//Parallel Distributed Processing:Foundations,Vol. 1. Cambridge,MA,USA:MIT Press,1986:194-281.
[12] HINTON G E,SALAKHUTDINOV R R. Reducing the dimensionality of data with neural networks[J]. Science,2006,331(5786):504-507.
[13] 王立中,管声启. 基于深度学习算法的带钢表面缺陷识别[J]. 西安工程大学学报,2017,31(5):669-674. WANG Lizhong,GUAN Shengqi. Strip steel surface defect recognition based on deep learning[J]. Journal of Xi'an Polytechnic University,2017,31(5):669-674.
[14] TOWELL G G,SHAVLIK J W. Knowledge-based artificial neural networks[J]. Artificial Intelligence,1994,70(1-2):119-165.
[15] 孙晨,周志华,陈兆乾. 神经网络规则抽取研究[J]. 计算机应用研究,2000,17(2):34-37. SUN Chen,ZHOU Zhihua,CHEN Zhaoqian. Study on rule extraction of neural network[J]. Application Research of Computers,2000,17(2):34-37.
[16] PENNING H L H D,AVILA GARCEZ A S D,LAMB L C,et al. A neural-symbolic cognitive agent for online learning and reasoning[J]. Proc. IJCAI,2011:1653-1658.
[17] TRAN S N,GARCEZ A S D. Knowledge extraction from deep belief networks for images[J]. Proc. IJCAI-Workshop Neural-Symbolic Learn. Reason.,2013:1-6.
[18] LI Shen,XU Hengru,LU Zhengdong. Generalize symbolic knowledge with neural rule engine[J]. 2018,arXiv preprint:1808.10326.
[19] 陈果,宋兰琪,陈立波. 基于神经网络规则提取的航空发动机磨损故障诊断知识获取[J]. 航空动力学报,2008(12):2170-2176. CHEN Guo,SONG Lanqi,CHEN Libo. Knowledge acquisition for aero2engine wear fault diagnosis based on rule extraction from neural networks[J]. Journal of Aerospace Power,2008(12):2170-2176.
[20] ZHENG Ni,ZHANG Lin,WANG Wenfeng,et al. Research on fault diagnosis method based on rule base neural network[J]. Journal of Control Science and Engineering,2017,2017:1-7.
[21] YU Jianbo,XI Lifeng,ZHOU Xiaojun. Intelligent monitoring and diagnosis of manufacturing processes using an integrated approach of KBANN and GA[J]. Computers in Industry,2008,59(5):489-501.
[22] HINTON G E. Training products of experts by minimizing contrastive Divergence[J]. Neural Computation,2002:1771-1800.
[23] PINKAS G. Reasoning,nonmonotonicity and learning in connectionist networks that capture propositional knowledge[J]. Artificial. Intelligence.,1995,77(2):203-247.
[24] TRAN S N,GARCEZ A S D. Deep logic networks:Inserting and extracting knowledge from deep belief networks[J]. IEEE Transactions on Neural Networks & Learning Systems,2018,29(2):246-258.
[25] HOLLAND J H. An introductory analysis with applications to biology[M]. Ann Arbor:University of Michigan Press,1975.
[26] FREITAS A A,CONCEICAO R I. A survey of evolutionary algorithms for data mining and knowledge discovery[M]. Ghosh A,Tsutsui S eds. Advances in Evolutionary Computing. Natural Computing Series. Springer,Berlin,Heidelberg,2003:819-845
[27] FERNANDO S O,AMY B. INSS:A hybrid system for constructive machine learning[J]. Neurocomputing,1999,28(1-3):191-205.
[28] MURPHY PM,AHA DW. UCI repository of machine learning databases[J]. Dept. Inf. Computation. Sci.,Univ. California,Irvine,CA,1994.
[29] KOHAVI R. A study of cross-validation and bootstrap for accuracy estimation and model selection[C]//Proceeding of the Fourteenth International Joint Conference on Artificial Intelligence. San Francisco,CA:Morgan Kaufmann,1995:1137-1143.
[30] BROSHEER B C,How smooth is smooth? Part-I,specification and evaluation of machined finishes[J]. American Machinist,McGrow-Hill Publishing Company,1948:97-112.
[31] LOU S J. Development of four in-process surface recognition systems to predict surface roughness in end milling[C]//Retrospective Theses and Dissertations,1997,Iowa State University. Ames. Lowa:Digital Repository@Iowa State University,http://lib.dr.iastate.edu/.