摘要
气门间隙异常是柴油机常见机械故障之一,对其进行准确的诊断对提高柴油机的使用寿命具有积极的作用.针对柴油机气门间隙异常的问题,在某直列6缸柴油机上模拟了不同气门故障,提出了基于双谱估计、图像处理以及分形理论相结合的故障诊断方法.该方法首先利用双谱估计对非线性、非高斯信号的敏感性质,分析了不同故障状态下振动信号中非高斯成分及二次相位耦合特性,然后通过图像处理技术将双谱图表示为以像素位置及对应颜色强度构成的三维空间曲面,最后利用分形理论提取该曲面的分形盒维数作为故障特征.结果表明:不同状态下柴油机振动信号的双谱及其图像分形维数明显可分,正常状态下的双谱峰值分布最为复杂、分形维数最大,故障状态下的分形维数分别处在不同的范围.因此,以振动信号的分形维数作为特征值可实现柴油机气门故障诊断.
Accurate diagnosis on valve clearance fault,which is one of common mechanical faults of diesel engine,has a positive effect on improving the service life of diesel engine.In this paper,a new approach based on bispectrum estimation,image processing and fractal theory was presented for the fault diagnosis of diesel engine valve train.Firstly,different valve clearance faults were simulated on the six-cylinder in-line diesel engine,and then vibration signals under different fault conditions were analyzed based on the sensitivity of bispectrum to nonlinear and nonGaussian signals.Secondly,image processing technology was used to make bispectrum to be the three-dimensional surface by pixel position and the corresponding color intensity with fractal characteristics.Finally,fractal theory was used to extract the surface fractal box dimension as the fault feature.Results show that bispectrum and its image fractal dimension are obvious different under different working conditions.In the normal state,bispectrum is the most complicated and fractal dimension is the largest.Other fractal dimensions of bispectrum are in different scopes. Therefore,using the vibration signal fractal dimension as the characteristic parameter can diagnose engine fault.
出处
《内燃机学报》
EI
CAS
CSCD
北大核心
2016年第3期274-280,共7页
Transactions of Csice
基金
国家科技支撑计划资助项目(2015BAF07B04)
关键词
柴油机
故障诊断
双谱
图像处理
分形维数
diesel engine
fault diagnosis
bispectrum
image processing
fractal dimension
