基于 RP与 NMF的内燃机气阀故障诊断方法

摘要

针对传统的内燃机故障振动诊断方法，难以有效提取故障特征，诊断精度较低的缺点，提出一种基于递归图（ recurrence plots，RP）与非负矩阵分解（ non－negative matrix factorization ，NMF）的内燃机故障诊断新方法。该方法是利用图像的方法来进行故障诊断：首先通过递归图将采集到的内燃机缸盖表面振动信号生成图像，然后用非负矩阵对得到递归图进行特征参数提取，最后用分类器进行分类识别完成故障诊断。将该方法应用于气阀机构8种工况下振动信号诊断实例中，结果表明：该方法克服了传统的振动诊断方法从时域或频域进行分析时参数选取和故障特征提取的难题，直接将信号生成图像，对图像进行自适应特征参数提取、分类识别，能有效诊断出内燃机气阀机构故障，故障识别精度高，为内燃机振动诊断探索了一条新途径。%In view of the traditional fault diagnosis method of internal combustion engine is difficult to extract the fault feature effectively and the fault diagnosis accuracy is low .A new fault diagnosis method based on recur-rence plots ( RP) and non-negative matrix factorization ( NMF) is put forward .The method is using image method to fault diagnosis:first use RP method to the collected cylinder head of internal combustion engine surface vibration signals, the images will be generated;and then use NMF method to the above obtained images for feature parameter extraction;finally the classifieris used for identification and classification to complete fault diagnosis .8 typical faults of diesel engine ’ s valve train were simulated .The vibration acceleration signals , which were acquired from the cylinder head , were analyzed with the proposed method .The results showed that IC Engine valve fault diagnosis method based on RP and NMF overcomes parameters selection and fault feature extraction problems of the tradition-al vibration diagnostic methods , directly to generate RP images , the adaptively to extract feature parameters of ima-ges.IC Engine valve fault diagnosis method based on RP and NMF can effectively diagnose of valve mechanism fault of internal combustion engine , fault recognition accuracy is high , explored the a new way for vibration diagno-sis of internal combustion engine .