基于K-S检验的变载荷齿轮故障诊断  

Using Kolmogorov-Smirnov Test on Diagnosis of Gear Faults under Fluctuating Load Conditions

作　　者：王细洋[1] 万在红[1] 

机构地区：[1]南昌航空大学,南昌330063

出　　处：《中国机械工程》2009年第9期1048-1052,1056,共6页China Mechanical Engineering

基　　金：航空基础科学基金资助项目(2006ZE56007)

摘　　要：齿轮载荷的变化会引起振动信号的变化,因此传统方法对变载荷齿轮箱的故障诊断不一定有效。将齿轮振动残余信号作为分析对象,以排除载荷变化的影响。对正常齿轮振动残余信号和待处理齿轮振动残余信号进行两样本Kolmogorov-Smirnov检验,获得两信号的统计距离和相似概率,并将统计距离和相似概率作为齿轮破损特征指标量,该特征指标量能较早地指示轮齿破损的发生。齿轮全生命周期运行的破损试验表明了这一特征指标量的有效性。Varying load can cause changes in a measured gearbox vibration signals, however, conventional techniques for fault diagnosis are based on the assumption that changes in vibration signals are only caused by deterioration of the gearbox. Gear motion residual signals, which represent the departure of time synchronous averaging signals from the average tooth--meshing vibration, were analyzed as source data due to its lower sensitiveness to the alternating load conditions. The gear motion residual signals in gear healthy state were used as reference sample,and the residual signals in unknown gear states were used as objective sample. A two--sample Kolmogorov--Smirnov test was then performed. The statistical distance of the test was used as a fault parameter which was capable of diagnosing the gear crack effectively with a higher level of confidence. The other statistical measures such as kurtosis, variance were also analyzed and computed. The performance of the technique was demonstrated using a full lifetime gear shaft vibration data history.

关 键 词：齿轮振动 故障诊断 Kolmogorov--Smirnov检验 变载荷 信号处理 

分 类 号：TH165.3[机械工程—机械制造及自动化]