短时AR谱分析及在内燃机故障诊断中的应用

针对利用传统短时Fourier变换(STFT)进行时频分析时不可能同时得到任意高时域分辨率和频域分辨率的问题,提出对传统的短时Fourier变换进行改进,在短时Fourier变换的计算中,利用自回归(AR)谱估计代替离散Fourier变换(DFT),得到了一种基于AR谱估计的短时AR谱分析方法。应用该方法对内燃机气阀机构的故障进行了分析,结果表明短时AR谱分析的估计性能大大改善,能够较好地给出信号的时频分布表示。

Short-Time AR Spectral Analysis and Its Application to Fault Diagnosis of Internal Combustion Engine

It's impossible to gain infinite high resolving power of time and frequency at the same time when short-time Fourier transform(STFT) is applied to time-frequency analysis.An alteration of STFT is put forward in this paper.In the calculation of STFT,discrete Fourier transform(DFT) is replaced by autoregressive(AR) spectral estimation,and a new method based on AR spectral estimation is obtained,namely short-time AR spectral analysis.The method is applied to the fault diagnosis of valve trains,and the results show that its estimation performance is improved greatly.It can give a better time-frequency distribution expression of the signals.