基于时频谱图与图像分割的柴油机故障诊断  

Diesel Engine Fault Diagnosis Based on Time-Frequency Spectrum Image and Image Segmentation

作　　者：蔡艳平[1] 李艾华[1] 王涛[1] 胡重庆[1] 

机构地区：[1]第二炮兵工程学院五系,陕西西安710025

出　　处：《内燃机学报》2011年第2期181-186,共6页Transactions of Csice

基　　金：博士后科学基金资助项目(20070411116);学院创新性基础研究基金资助项目(XY2009JJB33)

摘　　要：将图像分割理论引入柴油机故障诊断中,提出一种基于时频谱图、图像分割和模糊模式识别的柴油机故障诊断新方法.首先利用二进小波对柴油机缸盖振动信号进行预处理,然后用时频谱图对柴油机气门机构4种状态下的缸盖表面振动信号进行时频分析,并将谱图结果根据图像分割理论对其等高图进行分割,最后通过选取分割后图像的特征体质心位置、特征体面积、数目和熵作为特征参数,并利用模糊C均值聚类对图像进行分类识别.试验结果表明,新方法提取的振动信号图像几何特征与形状特征参数能充分反映柴油机气门工作状态的信息,对不同类型的气门故障均能正确诊断.The image segmentation theory is introduced into the diesel engine fault diagnosis.A new diesel fault diagnosis method is proposed by using the time-frequency spectrum image,image segmentation technology and fuzzy pattern recognition.Firstly,cylinder head vibration signal is pre-processed by the binary wavelet decomposition.Secondly,time-frequency spectrum image is generated from the vibration signals of valve in four different states.Based on theory of image segmentation,the geometrical and shape parameters are selected as the pattern classification feature.These feature parameters include the biggest characteristic body heart position,area,number and entropy lastly.The fault diagnosis on diesel valve gap abnormal fault is realized by fuzzy C-means pattern recognition.To examine the practicability and feasibility of the new method,the impact of noise on the proposed method is discussed.Experimental results show that the proposed method can segment the time-frequency image of different feature of vibra-tion acceleration signal.It has high identify ratio by using image geometric features and shape features as parameters to diagnose diesel engine valve fault.The time-frequency spectrum image geometric features and shape features can fully reveal the information of diesel engine valve working condition.

关 键 词：柴油机 故障诊断 时频谱图 图像分割 模式识别 

分 类 号：TK421.6[动力工程及工程热物理—动力机械及工程]