基于数据驱动的矿用卡车驱动系统智能故障诊断研究.pdf

国家重点研发计划（2019YFB2006400） 工程硕士专业学位论文 基于数据驱动的矿用卡车驱动系统智能故障 诊断研究 Research on Intelligent Fault Diagnosis of Mining Truck Drive System Based on Data Drive 作 者何 慧 导 师杨 真 教授 中国矿业大学 二〇二一年五月 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 学校代码 10290 UDC 密 级 公开 中国矿业大学 工程硕士专业学位论文 基于数据驱动的矿用卡车驱动系统智能故 障诊断研究 Research on Intelligent Fault Diagnosis of Mining Truck Drive System Based on Data Drive 作 者 何 慧 导 师 杨 真 申请学位 工程硕士专业学位 培养单位 矿业工程学院 学科专业 矿业工程 研究方向 智能开采 答辩委员会主席 姚强岭 评 阅 人 盲评 二〇二一年五月 万方数据 致谢致谢 时光飞逝，不知不觉我的硕士生涯已接近尾声，回想这两年的点点滴滴，酸 甜苦辣皆是珍贵的回忆，留下成长和收获的足迹。感谢陪我度过这段美好时光的 每一位值得尊敬的老师和每一位可爱的同学们，因为有你们的帮助和陪伴， 让我 掌握知识、解决问题、克服困难、收获成长，丰富了我这两年的学习和生活。 首先感谢我的导师杨真教授，本篇论文是在杨真教授的指导下顺利完成的， 感谢杨真教授对我的悉心指导和谆谆教诲，从论文的选题、项目的进行和论文最 终定稿， 杨真教授一直耐心的指导我， 在生活方面也是悉心关怀， 谢谢杨真教授。 杨真教授以渊博的学识、开阔的视野、严谨的态度、教书育人的精神、精益求精 的工作作风，深深激励和感染着我，让我由衷地敬佩，在此向杨真教授致以衷心 的感谢和崇高的敬意。 其次感谢此次国重项目团队对我学习和工作上的指导和帮助， 感谢中国矿业 大学机电学院苗运江老师和计算机学院张博老师对我的悉心指导， 同时感谢徐州 徐工矿业机械有限公司谢所、顾部、张工、房工、鹿工等领导对我的耐心帮助， 让我在学习科学知识的同时提高了自己的实践能力和创新能力。 再次感谢课题组的师兄们在学习和生活中给予我的帮助和支持， 感谢他们与 我一同分享青春的充实和快乐。在此特别感谢郭昌放师兄、姚文理师兄、武祥师 兄、 檀廷江师兄、 赵科师兄以及朱亚坤师兄给我的关心和帮助， 他们耐心指导我， 生活上对我也是关怀备至， 同时也感谢同级小伙伴张国华以及我亲爱的室友田苗 苗，我们一起学习互帮互助一起进步，给研究生生涯留下了美好的回忆。总之， 感谢实验室的所有人，感谢他们不辞劳苦不计回报地为我提供无私的帮助，没有 他们的帮助，我的论文无法顺利完成。 最后，我要向百忙之中参与审阅、评议本论文各位老师、向参与本人论文答 辩的各位老师表示由衷的感谢在矿大的这两年，因为有你们的关心和帮助，让 我倍感幸福，这段时光值得我回味一生。接下来的日子我会始终怀着感恩的心， 更加勤奋努力，奋勇直前，让自己变得更加优秀，更好的为社会做出贡献，我想 这是对所有栽培过我、帮助过我的人最好的回报吧。把最好的祝福献给你们，祝 你们身体健康、工作顺利。 万方数据 I 摘摘 要要 煤矿智能化技术是实现国家能源安全高效开采的核心技术。 矿用卡车作为煤 矿重要的采运设备，其重要部件出现故障可能会影响生产效率，甚至影响煤矿开 采安全。 因此，研究先进智能的矿用卡车故障诊断技术对于保障煤矿生产运输安 全高效地运行具有非常重要的研究意义。 驱动系统是矿用卡车的关键部件，本文以矿用卡车驱动系统为研究对象，采 用数据驱动的方式实现智能故障诊断。 首先分析了矿用卡车驱动系统的结构和故 障情况， 分别对轴承和齿轮的主要故障模式、 振动机理和振动特征频率进行分析， 为特征提取和智能故障诊断奠定了理论基础。 其次采用小波包进行去噪仿真分析 并进一步实现矿用卡车驱动系统振动信号的降噪处理； 采用集合经验模态分解对 振动信号进行分解，计算得出故障特征向量，实现故障特征提取。再次研究了基 于支持向量机的故障模式识别算法，为了提高故障状态识别精度，引入粒子群算 法、遗传算法及多种群遗传算法对支持向量机的惩罚因子 C 和核函数参数 g 进 行参数寻优， 通过四种模型的结果对比后构建了基于多种群遗传算法优化支持向 量机的矿用卡车驱动系统智能故障诊断模型。最后本文针对国内自主研发的 XDE320 电传动矿用自卸车驱动系统，开发了一套包含数据采集和存储模块、时 频分析模块、 特征提取模块和智能故障诊断模块这四个模块的矿用卡车智能故障 诊断系统，对采集到的振动数据进行时频分析，实现了该矿用卡车驱动系统的状 态监测和实时响应； 由于该矿卡的故障数据不足，所以通过已有的轴承和齿轮故 障数据集对信号进行故障特征提取和智能诊断， 验证了所设计的智能故障诊断系 统的实用性和可靠性； 随着矿用卡车运行数据的积累， 使系统能准确反应设备状 态并智能诊断出故障位置，提高设备的稳定性和维修的及时性，保障煤炭安全高 效生产。 论文中有图 67，表 18，参考文献 102 篇。 关键词关键词智能开采；矿用卡车；驱动系统；智能故障诊断；支持向量机 万方数据 II Abstract Intelligent coal mine technology is the core technology to realize the safe and efficient mining of national energy. As an important mining and transportation equipment for coal mines, the failure of important parts of mining trucks may affect production efficiency and even affect coal mining safety. Therefore, the study of advanced and intelligent fault diagnosis technology for mining trucks has very important research significance for ensuring the safe and efficient operation of coal mine production and transportation. The driving system is the key component of the mining truck. This article takes the mining truck driving system as the research object and adopts a data-driven approach to realize intelligent fault diagnosis. First, the structure and failure conditions of the mining truck drive system are analyzed, and the main failure modes, vibration mechanism and vibration characteristic frequencies of bearings and gears are analyzed respectively, which lays a theoretical foundation for feature extraction and intelligent fault diagnosis. Secondly, the wavelet packet is used for denoising simulation analysis and further realize the noise reduction processing of the vibration signal of the driving system of the mining truck; although the empirical mode decomposition can effectively decompose the signal, it has end effects and modal aliasing problems, so the collective empirical model State decomposition decomposes the vibration signal, calculates the energy ratio of the six eigenmode functions that best represent the state characteristics, and uses it as a fault feature vector to achieve fault feature extraction. The fault pattern recognition algorithm based on support vector machine is studied again. In order to improve the accuracy of fault state recognition, particle swarm optimization, genetic algorithm and multi-population genetic algorithm are introduced to optimize the parameters of the penalty factor C and the kernel function parameter g of the support vector machine. Through the comparison of the results of the four models, it is concluded that the support vector machine model optimized based on the multiple population genetic algorithm has the highest accuracy and the best perance. Thus, an intelligent fault diagnosis of the mining truck driving system based on the multiple population genetic algorithm optimization support vector machine is constructed. model. Finally, this article aims at the domestic self-developed XDE320 electric drive mining dump truck drive system. Through hardware equipment such as sensors, collectors, computers, and software such as LabVIEW and MATLAB, a set of data acquisition and storage modules, time-frequency analysis modules, The feature 万方数据 III extraction module and the intelligent fault diagnosis module are the four modules of the mining truck intelligent fault diagnosis system, which pers time-frequency analysis on the collected vibration data, and realizes the status monitoring and real-time response of the mining truck driving system; The fault data of the mining card is insufficient, so the fault feature extraction and intelligent diagnosis of the signal through the existing bearing and gear fault data set verify the practicability and reliability of the designed intelligent fault diagnosis system; as the mining truck runs The accumulation of data enables the system to accurately reflect the status of the equipment and intelligently diagnose the fault location, improve the stability of the equipment and the timeliness of maintenance, and ensure the safe and efficient production of coal. There are Figure 67, Table 18, and 102 references in the paper. Keywords smart mining; mining truck; wheel drive system; intelligent fault diagnosis; support vector machine 万方数据 IV 目目 录录 摘要摘要 ............................................................................................................................... I 目录目录 ............................................................................................................................ IV 图清单图清单 ..................................................................................................................... VIII 表清单表清单 ....................................................................................................................... XII 变量注释表变量注释表 ............................................................................................................. XIII 1 1 绪论绪论 ........................................................................................................................ 1 1 1.1 研究背景及意义.................................................. 1 1.2 国内外研究现状.................................................. 2 1.3 研究内容........................................................ 9 1.4 研究方法与技术路线............................................. 10 2 2 矿用卡车驱动系统故障机理及振动特征分析矿用卡车驱动系统故障机理及振动特征分析 .................................................. 1111 2.1 矿用卡车驱动系统结构和故障分析................................. 11 2.2 轴承故障模式与机理分析......................................... 13 2.3 齿轮故障模式与机理分析......................................... 16 2.4 本章小结....................................................... 18 3 3 矿用卡车驱动系统的故障特征提取矿用卡车驱动系统的故障特征提取 ................................................................. . 1919 3.1 小波分析....................................................... 19 3.2 小波包分析..................................................... 21 3.3 信号的小波包消躁方法........................................... 23 3.4 EMD 基本理论 ................................................... 28 3.5 EEMD（Ensemble Empirical Mode Decomposition）.................. 31 3.6 振动信号的故障特征提取......................................... 34 3.7 本章小结....................................................... 40 4 4 矿用卡车驱动系统智能故障诊断矿用卡车驱动系统智能故障诊断 ...................................................................... 4242 4.1 支持向量机..................................................... 42 4.2 粒子群算法..................................................... 44 4.3 遗传算法....................................................... 46 4.4 支持向量机参数优化............................................. 52 4.5 智能故障诊断................................................... 55 4.6 本章小结....................................................... 58 万方数据 V 5 5 矿用卡车矿用卡车驱动系统智能故障诊断系统设计驱动系统智能故障诊断系统设计 ...................................................... 5959 5.1 开发平台简介................................................... 59 5.2 信号采集和存储模块............................................. 59 5.3 时频分析模块................................................... 66 5.4 故障特征提取模块............................................... 72 5.5 智能故障诊断模块............................................... 74 5.6 本章小结....................................................... 76 6 6 结论与展望结论与展望 .......................................................................................................... 7777 6.1 结论........................................................... 77 6.2 展望........................................................... 77 参考文献参考文献 ..................................................................................................................... 79 作者简历作者简历 ..................................................................................................................... 85 学位论文原创性声明学位论文原创性声明 ................................................................................................. 86 学位论文数据集学位论文数据集 ......................................................................................................... 87 万方数据 VI Contents Abstract ........................................................................................................................ II Contents .................................................................................................................... IVI 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 at Home and Abroad ...................................................................... 2 1.3 Research Content ..................................................................................................... 9 1.4 Research s and Technical Routes ............................................................... 10 2 Analysis of Failure Mechanism and Vibration Characteristics of Driving System of Mining Truck ......................................................................................................... 11 2.1 Structure and Failure Analysis of Driving System of Mining Truck ..................... 11 2.2 Analysis of Bearing Failure Mode and Mechanism ............................................... 13 2.3 Gear Failure Mode and Mechanism Analysis ........................................................ 16 2.4 Chapter Summary .................................................................................................. 18 3 Extraction of Fault Features of Mining Truck Drive System ............................. 19 3.1 Wavelet Analysis .................................................................................................... 19 3.2 Wavelet Packet Analysis ........................................................................................ 21 3.3 Signals Wavelet Packet De-noising ......................................................... 23 3.4 EMD Basic Theory ................................................................................................ 28 3.5 EEMD（Ensemble Empirical Mode Decomposition） ........................................ 31 3.6 Fault Feature Extraction of Vibration Signal ......................................................... 34 3.7 Chapter Summary .................................................................................................. 40 4 Intelligent Fault Diagnosis of Driving System of Mining Truck ......................... 42 4.1 Support Vector Machines ....................................................................................... 42 4.2 Particle Swarm Algorithm...................................................................................... 44 4.3 Genetic Algorithm .................................................................................................. 46 4.4 Support Vector Machine Parameter Optimization ................................................. 52 4.5 Intelligent Fault Diagnosis ..................................................................................... 55 4.6 Chapter Summary .................................................................................................. 58 5 Design of Intelligent Fault Diagnosis System for Driving System of Mine Truck 万方数据 VII ...................................................................................................................................... 59 5.1 Introduction to Development Plat .................................................................. 59 5.2 Signal Acquisition and Storage Module ................................................................ 59 5.3 Time-frequency Analysis Module .......................................................................... 66 5.4 Fault Feature Extraction Module ........................................................................... 72 5.5 Intelligent Fault Diagnosis Module ....................................................................... 74 5.6 Chapter Summary .................................................................................................. 76 6 Conclusion and Outlook ......................................................................................... 77 6.1 Conclusion .