齿轮故障诊断技术研究

Research on Gear Fault Diagnosis Technology

齿轮故障诊断技术的本质是通过观察及分析齿轮不同状态理化性质变化,达到判别齿轮故障的效果。其研究目的是实现齿轮故障早发现、准诊断,提升各型设备的加工水平及使用寿命。传统的油样分析技术通过设备用润滑油的磨屑情况分析,判断齿轮工作状态。其优势是检出限低,可以把握微米级磨损变化。但是其不具备实时性,且需要人为观察,判别较为繁琐。采用齿轮振动信号进行故障诊断是目前较为常见的另一种诊断方法。通过加速度传感器对齿轮振动信号进行采集,在利用包括传统时域及频域分析在内的多种分析手段,对齿轮故障情况进行判别。其具备不开箱、成本低、时实性强等优点,但其检出限较高。分析基于振动信号的故障诊断办法,在介绍常见时域及频域分析基础上,引出基于特向选择、特向优化且与分类函数结合的先进齿轮故障判别方法。基于稀疏滤波的特向优化技术是目前基于振动信号进行齿轮故障诊断的发展方向。

The essence of gear fault diagnosis technology is to identify gear faults by observing and analyzing the changes of physical and chemical properties of gears in different states.The purpose of study is to realize early detection and accurate diagnosis of gear faults and improve the processing level and service life of various types of equipment.Traditional oil sample analysis technology judges gear working status by analyzing the wear debris in lubricating oil used in equipment.Its advantage is low detection limit,and it can distinguish the change of micron-grade wear,but not in real-time,and requires human observation and cumbersome discrimination.Fault diagnosis based on gear vibration signal is another common diagnostic method at present.Gear vibration signals are collected by acceleration sensor,and then analyzed by a variety of analysis methods including traditional time-domain analysis and frequency-domain analysis to distinguish gear fault.It has the characteristics of no unpacking,low cost and real-time processing,but its detection limit is high.In this paper,the method of fault diagnosis based on vibration signals was analyzed in detail.On the basis of introducing common time-domain analysis and frequency-domain analysis,an advanced gear fault diagnosis method based on special selection,special optimization and in combination with classification function was proposed.The special optimization technology based on sparse filtering was the development direction of gear fault diagnosis based on vibration signals.