基于缸盖振动信号包络线相关分析的柴油机故障预警方法研究

针对柴油机故障特征易受到环境、载荷等因素的干扰，从而导致许多传统意义上的故障特征参数不再敏感的问题，提出了一种基于缸盖振动信号包络线相关分析的柴油机故障预警方法，对一台6缸柴油机的缸盖振动信号对比了希尔伯特变换和n阶极值插值两种方法的包络线提取效果，并对其包络线进行相关分析，得到了相似系数、距离标准差以及余弦相似度等相关分析结果。研究结果表明：在传统特征参数不敏感的情况下，相关分析可以更好地反映机组的故障状态。本文研究结果可为实现柴油机故障的早期预警奠定基础。

A diesel engine fault early warning method based on correlation analysis of the cylinder head vibration signal envelope line

The fault characteristics of diesel engines are easily affected by factors such as environment and load, resulting in many traditional fault feature parameters being no longer sensitive. In this work, a diesel engine fault warning method based on correlation analysis is proposed which focuses on the cylinder head vibration signal envelope line. Two different methods of extracting the vibration signal envelope lines of the cylinder head in a 6-cylinder diesel engine-Hilbert transform and n-order extremum interpolation-are compared. Based on the analysis of the envelope lines, the correlation coefficients, distance standard deviation and cosine similarity are obtained. The results show that the fault state of the diesel engine can be better illustrated by correlation analysis in the case when the traditional characteristic parameters are not sensitive. The correlation analysis thus lays a foundation for an early warning system for diesel engine faults.