应用循环自相关函数诊断齿轮故障

本文讨论机械信号的循环平稳特性，指出这种循环平稳性主要表现为二阶．推导了二阶循环统计量——循环自相关函数的解调方法，该方法通过搜索循环频率而将调制源分离在循环域的低频段，并且调幅和调频信号有不同的解调结果．应用于齿轮诊断，有效地提取出裂纹和断齿故障的特征．研究表明：将机械信号视为循环平稳信号，可以获得统计量的变化信息，这种信息在机械监测和诊断中可以充分利用．

Applying Cyclic Autocorrelation Function to Diagnose Gear Faults

The paper discusses the cyclostationarity that manifests second-order in mechanical signals. Based on the second-order statistics, cyclic autocorrelation function, a method of demodulation is deduced. By scanning the cyclic frequency domain, the modulatiors are separated and located at the lower frequency range. Furhtermore,the results both the signal with amplitude modulated and the one with frequency modulated are different. Applying the method to process the gear signals,the faults of crackle and broken in the eteeth are diagnosed. The research demonstrates that the considered mechanical signals as cyclostationary, the periodic time-varying information of statistics could be obtained and exploited to monitor and diagnose the machinery.