间歇性低速重载行星轮系故障演化机理及其定量诊断方法研究

Low speed and heavy load (LSHV) gearboxes are the essential transmission systems in modern industry which often operates under heavy load, impact, variable load and other complex tough conditions. Their key components are often subject to damage modes such as fatigue crack, spall and broken. This proposal takes the LSHV planetary gearbox as the research object and focuses on dynamic evolution mechanism and quantitative diagnosis fault feature extraction method of early injury faults. On the basis of rigid-flexible coupling multi body dynamics theory and by taking both centralized parameter method and dynamic extended finite element method as technical means, the multi-degree-of-freedom dynamic model of gearbox with typical micro-injury fault which can be extended will be built. Natural and vibration response characteristics and related experiments will be investigated under special working condition during fault propagation process. Characterization laws of fault feature information and nonlinear evolution mechanism in the fault development process will be deeply researched. According to this type equipment’s vibration signals characteristics (multi-resonance, multi-impacts, multi-modulation, multi-mode mixing and etc.), adaptive atom dictionary construction methods to reflect the fault severity degree will be explored. Adaptive sparse decomposition quantitative fault diagnosis theory and methods will be mainly researched. Thereby the quantitative diagnosis theory and technique of early fault diagnosis for LSHV gears transmission system will be provided to satisfy the requirement of economic development and obtain distinct economic and social benefit.

低速重载齿轮箱是现代工业设备的关键传动系统，常工作于冲击和变载荷等复杂工况下，其核心零部件易产生裂纹、剥落甚至断裂等损伤性故障。课题拟以低速重载行星齿轮箱为研究对象，深入研究其早期损伤性故障动态演化机理及其故障定量特征提取方法。以刚柔耦合动力学为理论基础，以集中参数法和动态小波扩展有限元分析为技术手段，建立齿轮箱非线性多自由度微损伤故障可扩展的动力学模型；研究特殊工况下行星齿轮传动系统故障动态发展过程中的固有特性和振动响应特性并进行相关实验验证；探索与故障发生发展过程相对应的故障扩展机理和特征及故障信息演变规律；针对此类设备振动信号特点（多共振、多冲击、多调制和多模式混淆等），探索反映故障损伤程度即定量信息特征的自适应基原子库构造方法，重点研究基于信号多特征的自适应稀疏分解定量诊断理论与方法，提供具有普遍意义低速重载齿轮箱故障定量诊断理论方法，满足国民经济发展的迫切需求。

本课题采用理论分析和数值仿真以及实验和工程实际数据分析等研究手段，重点研究行星轮系关键零部件轴承及齿轮的故障机理及其定量诊断理论与方法。. 针对轴承损伤性故障机理及其定量故障诊断方法方面，主要研究成果在于：建立了故障滚动轴承非线性动力学模型，分析了处于轴承主要承载区间内，且尺寸相差滚珠角间距自然倍数的轴承外圈缺陷对载荷分布、轴承刚度、前二阶固有频率以及振动加速度时域信号和频域信号的影响；提出了基于垂-平同步指标的滚动轴承外圈缺陷宽度定量诊断及定位诊断新方法；提出了基于级联字典匹配追踪的滚动轴承故障定量诊断方法；提出了基于定量映射模型的滚动轴承定量定位故障诊断方法；此外，进行了基于数据驱动的滚动轴承剩余使用寿命预测的初步研究。所提出的系列方法均在仿真信号、实验信号或工程实际中都取得了很好的效果。. 针对行星轮系故障机理及故障特征提取方法方面，其主要研究成果在于：研究了行星齿轮箱故障激励、冲击激励以传递误差等多重激励下的系统振动响应，建立了多自由度行星齿轮箱动力学模型；基于行星齿轮箱的几何关系，推导了齿轮副啮合相位和重合度的关系；提出了一种基于啮合振动特性的行星齿轮箱振动信号唯象模型；提出了基于集中参数模型及欧拉梁的振动理论联合仿真行星齿轮箱振动信号的方法；提出了基于改进VMD与快速谱峭度方法结合的行星齿轮箱故障诊断方法；提出了基于转域重采样和变多尺度形态滤波的行星齿轮箱故障特征提取方法。上述方法均采用了齿轮振动仿真信号和实验信号或工程信号验证了其可行性及有效性。. 项目期内，发表中英文期刊和会议论文39篇（SCI收录24篇，包括ESI高被引7篇，HOT热点论文2篇），申请国家发明专利12项，已获授权4项，申请国际PCT专利1项。本项目成果作为主要支撑内容之一，获2019教育部高校成果自然科学奖二等奖。