柴油机振动信号的分形特征及故障诊断  

Fractal Characteristics and Fault Diagnosis of Vibration Signal of Diesel Engines

作　　者：董安[1] 潘宏侠[1] 

机构地区：[1]中北大学机械工程与自动化学院,太原030051

出　　处：《噪声与振动控制》2014年第2期144-147,共4页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(50875247)

摘　　要：把小波包降噪与关联维数相结合用在柴油机故障诊断中,讨论了关联维数计算中相关参数的选取。首先对柴油机缸盖振动信号进行小波包降噪处理,通过自相关函数法求延迟时间,用饱和关联维数法确定最小嵌入维数,并用改进的G-P算法求关联维数。分析发现,小波包可以有效降低噪声干扰,柴油机在不同工况下具有不同的关联维数,该方法可有效应用于柴油机的故障诊断中。Application of the wavelet packet de-noising and correlation dimension in fault diagnosis of diesel engines is studied. Selection of related parameters in the calculation of the correlation dimension is discussed. First of all, the vibration signal of the cylinder-head of a diesel engine is processed using wavelet packet de-noising method. Then, the self-correlation function method is used for computation of the time delay. The saturation correlation dimension method is used to determine the minimum embedding dimension, and the improved G-P algorithm is employed to calculate the correlation dimension. The analysis shows that the wavelet packet de-noising method can effectively reduce the noise disturbance and the diesel engine has different correlation dimensions in different conditions. This method can be applied to the fault diagnosis of diesel engines.

关 键 词：振动与波 小波包降噪 关联维数 柴油机 故障诊断 

分 类 号：TK427[动力工程及工程热物理—动力机械及工程]