基于加权排列熵和ELM的通风机轴承故障诊断研究.pdf

硕士学位论文 基于加权排列熵和 ELM 的通风机轴承故障 诊断研究 Research on Fault Diagnosis of Ventilator Bearing Based on Weighted Permutation Entropy and ELM 作 者张旭 导 师吴新忠 副教授 中国矿业大学 二○二○年六月 万方数据 学位论文使用授权声明学位论文使用授权声明 本人完全了解中国矿业大学有关保留、使用学位论文的规定，同意本人所撰 写的学位论文的使用授权按照学校的管理规定处理 作为申请学位的条件之一， 学位论文著作权拥有者须授权所在学校拥有学位 论文的部分使用权，即①学校档案馆和图书馆有权保留学位论文的纸质版和电 子版，可以使用影印、缩印或扫描等复制手段保存和汇编学位论文；②为教学和 科研目的，学校档案馆和图书馆可以将公开的学位论文作为资料在档案馆、图书 馆等场所或在校园网上供校内师生阅读、浏览。另外，根据有关法规，同意中国 国家图书馆保存研究生学位论文。 （保密的学位论文在解密后适用本授权书） 。 作者签名 导师签名 年 月 日 年 月 日 万方数据 中图分类号 TP183 学校代码 10290 UDC 621.3 密 级 公开 中国矿业大学 硕士学位论文 基于加权排列熵和 ELM 的通风机轴承 故障诊断研究 Research on Fault Diagnosis of Ventilator Bearing Based on Weighted Permutation Entropy and ELM 作 者 张旭 导 师 吴新忠 副教授 申请学位 工学硕士学位 培养单位 信息与控制工程学院 学科专业 控制科学与工程 研究方向 信号处理；模式识别 答辩委员会主席 孙晓燕 教授 评 阅 人 二○二○年六月 万方数据 致谢致谢 三年硕士研究生生活即将结束，本论文是这一学习阶段研究成果的体现，值 此之际，向众多给予我帮助的人表达感谢。 感谢我的导师吴新忠老师。吴老师在研究过程中一步步耐心指引，帮助我将 理论应用到实际，不断提升我的专业能力，以丰富的专业知识与严谨的治学态度 时刻激励我，我将秉承“不忘初心”的态度踏实完成今后的工作学习。 感谢任子晖教授在学习与生活上对我的引导与帮助， 任老师踏实的科研精神 与严于律己的工作态度是我学习的楷模。感谢马小平教授对实验室的带领， 让我 有更多学习历练的机会。 感谢唐守锋老师和袁小平老师在论文的开题阶段对论文 的研究方向、研究内容的指导以及肯定。感谢师兄师姐朱静雅、陈炳光、陈明、 王翠、 倪婷婷、王琦、汪伟锋、袁猛、周成、夏令祥、张兆龙对我学习上的帮助； 感谢同届胡建豪、邱天舒时常与我进行学术交流，能一起走过硕士阶段是我荣幸 之至；感谢师弟师妹李博华、张芝超、陈昌、耿柯、董浩然、韩正化、陈泽彭、 蔡蔓利让我的硕士生活充满乐趣。 由衷感谢在百忙之中抽出时间参与论文评审、答辩的各位专家、教授，您们 提出的宝贵意见使我的研究更加完善。 万方数据 I 摘摘 要要 滚动轴承是煤矿通风机的重要零件， 对其进行运行状态监测是保证煤矿安全 生产的关键技术，本文将不同状态下，通风机轴承振动信号作为研究对象，以信 号特征参数为主要依据实现轴承故障预警及分类诊断。 在故障预警环节， 利用轴承处于不同运行状态时，振动信号的梅尔频率倒谱 系数存在差异这一特点，将其第一维参数与信号谱熵相乘得到谱熵梅尔积特征 MFPH， 提高信号特征的抗噪能力， 同时最大化正常信号与故障信号间的差异； 通过贝叶斯信息准则确定分类簇数，利用模糊 C 均值聚类算法自适应获取信号 谱熵梅尔积特征参数的高低门限，最后通过双门限法找到故障起始点，实现自适 应故障预警。 在信号分解环节，以经验模态分解EMD为基本分解方法，采用相关系数法 剔除结果中的虚假分量；采用支持向量回归机SVR延拓再分解的方法抑制端点 效应问题；使用自适应集总局部积分均值分解AELIMD算法，通过噪声数据辅 助分析缓解 EMD 的模态混叠现象，通过与 EMD、互补集总经验模态分解的对 比实验，证明应用 AELIMD 方法分解信号能得到更真实的频率成分。 在特征提取环节， 介绍了信号能量、排列熵及加权排列熵特征向量的构成方 法，进行基于真实滚动轴承振动信号的分析实验，证明能量向量因稳定性不好， 故不适合作为代表特征， 而相比于排列熵，加权排列熵特征向量同类中特征更统 一，不同类间差异更明显，因此将其确定为表征信号的特征向量。 在分类诊断环节，以极限学习机ELM作为故障分类的基础方法，用差分进 化DE算法对 ELM 输入权值及隐藏层阈值进行参数寻优；为使结果更逼近全局 最优解，加入模拟退火SA算法进行二次寻优；最后，进行轴承实际振动信号状 态分类实验，将 ELM、DE-ELM 及 SA-DE-ELM 模型作为分类器，并通过支持 向量机及其相应改进算法进行对比，证明在加权排列熵特征提取的条件下，应用 SA-DE-ELM 模型鉴别轴承故障类型能获得更高的准确率且诊断时间更短。 该论文有图 73 幅，表 14 个，参考文献 109 篇。 关键词关键词经验模态分解；谱熵梅尔积；加权排列熵；极限学习机；故障诊断 万方数据 II Abstract Monitoring running status of rolling bearing which is the vital component of ventilator is the key technology to ensure safe production of coal mine. In this paper, vibration signals of rolling bearing in different states are selected as study subject, and the bearing fault warning and classification diagnosis are based on signal characteristic parameters. In fault early warning step, mel frequency cepstrum coefficient of vibration signal is different in different operating states, and its first dimension parameter is multiplied with spectral entropy to obtain product of spectral entropy and MFCC0 MFPH, which improves noise resistance of signal characteristic and maximizes the difference between normal signal and fault signal. The number of classification clusters is decided by bayesian ination criterion, moreover, fuzzy C-means algorithm is used to adaptively obtain the high and low thresholds about MFPH of signal. Finally, double threshold is used to find the starting point of fault signal, which realizes fault early warning adaptively. In signal decomposition step, Empirical Mode Decomposition EMD is taken as basic decomposition , and correlation coefficient is used to eliminate false components. Support Vector Regression SVR extension followed by decomposition is adopted to suppress end effect. Adaptively Ensemble Local Integral Mean Decomposition AELIMD is used to alleviate mode mixing of EMD through assistance of noise data analysis., and it is proved that AELIMD can be used to decompose signals to obtain more realistic frequency components through comparison experiments with EMD and Complementary Ensemble Empirical Mode Decomposition. In feature extraction step, the composition s of feature vector about energy, permutation entropy, and weighted permutation entropy are introduced. It is proved that energy vector is not suitable as representative feature because of its poor stability through analysis and experiments based on actual rolling bearing vibration signals. Compared with permutation entropy, weighted permutation entropy feature vector has intra-class uniity and inter-class differences. Therefore, weighted permutation entropy feature vector is determined as eigenvector representing original signal. In classification and diagnosis step, Extreme Learning Machine ELM serves as the basic for fault classification, and differential evolution DE algorithm is used to per parameter optimization on weight and hidden layer threshold of 万方数据 III ELM. Simulated annealing SA algorithm is added for second optimization. ELM, DE- ELM, and SA-DE-ELM models are used as classifiers to classify experiments of actual bearing vibration signal states compared by SVM. It is proved that under the condition of weighted permutation entropy feature extraction, using SA-DE-ELM model to identify bearing fault types can obtain higher accuracy and shorter diagnosis time. This paper includes 73 figures, 14 tables and 109 references. Keywords Empirical mode decomposition; Product of spectral entropy and MFCC0; Weighted permutation entropy; Extreme learning machine; Fault diagnosis 万方数据 IV 目目 录录 摘要摘要 ............................................................................................................................ I I 目录目录 .......................................................................................................................... IVIV 图清单图清单 .................................................................................................................. VIIIVIII 表清单表清单 .................................................................................................................... XIIXII 变量注释表变量注释表 .......................................................................................................... XIIIXIII 1 1 绪论绪论 ........................................................................................................................ 1 1 1.1 研究背景意义.................................................... 1 1.2 故障诊断研究现状................................................ 1 1.3 论文主要研究内容及章节安排...................................... 8 2 2 通风机轴承故障机理分析通风机轴承故障机理分析 .................................................................................. 1010 2.1 煤矿通风机概述................................................. 10 2.2 滚动轴承的结构与振动原理....................................... 11 2.3 滚动轴承故障分析............................................... 13 2.4 本章小结....................................................... 14 3 3 通风机滚动轴承故障预警通风机滚动轴承故障预警 .................................................................................. 1515 3.1 轴承振动信号的谱熵梅尔积特征参数............................... 15 3.2 轴承振动信号故障起始点自适应检测............................... 18 3.3 基于谱熵梅尔积的滚动轴承故障预警............................... 21 3.4 本章小结....................................................... 25 4 4 基于基于 EMDEMD 的通风机轴承振动信号分解的通风机轴承振动信号分解 .............................................................. 2626 4.1 希尔波特黄变换................................................. 26 4.2 EMD 的端点效应 ................................................. 31 4.3 EMD 的模态混叠 ................................................. 41 4.4 本章小结....................................................... 48 5 5 基于加权排列熵的通风机轴承故障特征提取基于加权排列熵的通风机轴承故障特征提取 .................................................. 5050 5.1 故障信号特征分析............................................... 50 5.2 IMF 的特征信息 ................................................. 51 5.3 基于 AELIMD 的轴承故障特征提取.................................. 55 5.4 实际轴承振动信号特征提取及对比分析............................. 57 5.5 本章小结....................................................... 65 万方数据 V 6 6 基于基于 ELMELM 的通风机轴承故障诊断的通风机轴承故障诊断 ...................................................................... 6666 6.1 极限学习机原理................................................. 66 6.2 基于差分进化的极限学习机优化改进............................... 70 6.3 基于 DE-SA-ELM 的通风机轴承故障诊断............................. 75 6.4 本章小结....................................................... 81 7 7 总结与展望总结与展望 .......................................................................................................... 8282 7.1 总结........................................................... 82 7.2 展望........................................................... 83 参考文献参考文献 .................................................................................................................. 8484 作者简历作者简历 .................................................................................................................. 9292 学位论文原创性声明学位论文原创性声明 .............................................................................................. 9393 学位论文数据集学位论文数据集 ...................................................................................................... 9494 万方数据 VI Contents Abstract ........................................................................................................................ II Contents ..................................................................................................................... VI List of Figures ......................................................................................................... VIII List of Tables ............................................................................................................. XII List of Variables...................................................................................................... XIII 1 Introduction ............................................................................................................... 1 1.1 Research Background and Significance ................................................................... 1 1.2 Research Status of Fault Diagnosis.......................................................................... 1 1.3 Main Research Contents and Arrangements for the Chapter ................................... 8 2 Failure Mechanism Analysis of Ventilator Bearing ............................................. 10 2.1 Introduction of Coal Mine Ventilator ..................................................................... 10 2.2 Structure and Vibration Principle of Rolling Bearing ............................................ 11 2.3 Failure Analysis of Rolling Bearing ...................................................................... 13 2.4 Chapter Summary .................................................................................................. 14 3 Fault Warning of Ventilator Rolling Bearing ....................................................... 15 3.1 MFPH Characteristic Parameters of Bearing Vibration Signal ............................. 15 3.2 Failure Starting Point Adaptive Detection of Bearing Vibration Signal ................ 18 3.3 Fault Warning of Rolling Bearing Based on MFPH .............................................. 21 3.4 Chapter Summary .................................................................................................. 25 4 Vibration Signal Decomposition of Ventilator Bearing Based on EMD ............ 26 4.1 Hilbert-Huang Trans....................................................................................... 26 4.2 End Effect of EMD ................................................................................................ 31 4.3 Mode Mixing of EMD ........................................................................................... 41 4.4 Chapter Summary .................................................................................................. 48 5 Feature Extraction of Ventilator Bearing Based on Weighted Permutation Entropy ....................................................................................................................... 50 5.1 Characteristic Analysis of Fault Signal .................................................................. 50 5.2 Characteristic Ination of IMF ......................................................................... 51 5.3 Feature Extraction of Bearing Faults Based on AELIMD ..................................... 55 5.4 Feature Extraction and Comparative Analysis of Actual Bearing Vibration Signals 万方数据 VII ...................................................................................................................................... 57 5.5 Chapter Summary .................................................................................................. 65 6 Ventilator Bearing Fault Diagnosis Based on ELM ............................................. 66 6.1 Principle of Extreme Learning Machine ................................................................ 66 6.2 Optimization and Improvement of ELM Based on Differential Evolution ........... 70 6.3 Ventilator Bearing Fault Diagnosis Based on DE-SA-ELM .................................. 75 6.4 Chapter Summary .................................................................................................. 81 7 Summary and Outlook ........................................................................................... 82 7.1 Summary ................................................................................................................ 82 7.2 Outlook .................................................................................................................. 83 References ................................................................................................................... 84 Author’s Resume ........................................................................................................ 92 Declaration of Thesis Originality ............................................................................. 93 Thesis Data Collection ............................................................................................... 94 万方数据 VIII 图清单图清单 图序号 图名称 页码 图 1-1 故障预警及诊断流程图 3 Figure 1-1 Flow chart of fault warning and diagnosis 3 图 2-1 轴流式通风系统 10 Figure 2-1 Axial ventilation system 10 图 2-2 深沟球轴承结构图 11 Figure 2-2 Structure diagram of deep groove ball bearing 11 图 2-3 滚动轴承振动原理 11 Figure 2-3 Principle of rolling bearing vibration 11 图 2-4 滚动轴承故障类型 14 Figure 2-4 Fault types of rolling bearing 14 图 3-1 分帧原理 16 Figure 3-1 Principle of framing 16 图 3-2 MFCC 计算流程 17 Figure 3-2 MFCC calculation process 17 图 3-3 双门限法原理 20 Figure 3-3 Principle of doubl