基于变分模态分解和交叉小波变换的柴油机失火故障诊断

针对强噪声干扰下柴油机失火故障难以诊断的问题,提出一种基于变分模态分解(variational mode decomposition,VMD)与交叉小波变换(cross wavelet transform,XWT)的柴油机失火故障诊断方法。该方法首先通过VMD将缸盖振动信号进行分解、自适应消噪及信号重构,再利用XWT对任意两个连续工作循环信号进行时频相关分析,进一步消除振动信号中的干扰噪声以提取柴油机燃烧特征,最后通过计算时频空间各缸能量占比进行柴油机失火故障诊断。通过对仿真信号分析及柴油机失火故障诊断,结果表明:该方法可以消除强噪声干扰,提取柴油机燃烧周期瞬态振动冲击特征,有效地识别柴油机失火故障。

Fault Diagnosis of Diesel Engine Misfire Based on VMD and XWT

Aiming at the problem that diesel engine misfire is difficult to diagnose under strong noise interference, a diesel engine misfire fault diagnosis method based on Variational Mode Decomposition (VMD) and Cross Wavelet Transform (XWT) is proposed. Firstly, cylinder head vibration signal is segmented according to the diesel engine working cycle. Secondly, each working cycle vibration signal is resampled on the equally angle. And then, the cylinder head vibration signal is decomposed, denoised and reconstructed by VMD. Further, the time-frequency correlation analysis of any two consecutive working cycle signals is carried out by using cross-wavelet transform (XWT) to eliminate the non-periodic components and random interference noise in the vibration signal to extract the diesel combustion features. Finally, the diesel engine misfire fault diagnosis is performed by calculating the energy ratio of each cylinder in the time-frequency space. The simulation signal and actual signal analysis show that this method can eliminate the strong noise interference, extract the periodic transient shock features of the diesel engine combustion, and effectively identify the diesel engine misfire fault cylinder.