基于同步压缩-交叉小波变换算法的齿轮故障诊断研究

Research on Gear Fault Diagnosis Based on Synchronous Compression Cross Wavelet Transform

齿轮超负荷使用或者在不良环境下使用时,容易造成齿轮失效;齿轮的失效形式主要有断齿故障、齿面点蚀、齿轮磨损等,此外齿轮制造过程中也存在固有误差,传统的齿轮故障诊断通常使用振动加速度传感器或者SCADA数据进行处理,但振动加速度传感器与SCADA通常价格昂贵,且会有大量的数据冗余,不便于后期信号处理;本设计拟采用一种同步压缩-交叉小波变换算法,在齿轮故障机理分析的基础上,设计了故障诊断实验装置,对正常齿轮、断齿和磨损情况下的故障特性进行提取,从而对故障进行准确诊断,经实验验证,该方法在齿轮断齿检测上准确率达100%,对齿轮磨损检测达到98%以上,对齿面点蚀诊断准确率较低,本研究中暂未涉及,后续研究主要集中在对算法改进,提高对齿面点蚀的准确判断。

When the gear is used under overload or under bad environment,it is easy to cause gear failure.The failure forms of gears mainly include broken teeth fault,tooth surface pitting,gear wear,etc.in addition,there are inherent errors in the gear manufacturing process.The traditional gear fault diagnosis usually uses vibration acceleration sensor or SCADA data for processing,but the vibration acceleration sensor and SCADA are usually expensive,and there will be a lot of data redundancy,which is not convenient for later signal processing.In this design,a synchronous compression Cross Wavelet Transform algorithm is proposed.Based on the analysis of gear fault mechanism,a fault diagnosis experimental device is designed to extract the fault characteristics of normal gear,broken gear and wear conditions,so as to accurately diagnose the fault.The experimental verification shows that the accuracy of this method is 100%in gear fault detection and 98%in gear wear detection,the accuracy of the diagnosis of tooth surface pitting is low,which is not involved in this study.The follow-up research focuses on the improvement of algorithm to improve the accuracy of the judgment of tooth surface pitting.