摘要
针对柴油机振动信号的非平稳特性和在现实条件下难以获得大量故障样本的实际情况,提出了一种经验模式分解和支持向量机相结合的故障诊断方法。运用经验模式分解方法对气阀机构不同工况下的去噪缸盖振动信号进行分析,计算各固有模式函数的方差贡献率以确定包含故障特征信息的主要成分,对求得的各固有模式函数分别计算其能量矩,并将能量矩作为支持向量机的输入特征向量,以判断柴油机的工作状态和故障类型。试验结果表明:该方法在小样本情况下也具有较高的精度和较强的泛化能力,但转速不同时需重新采样以保证足够的诊断精度。
According to the non-stationarity characteristics of the vibration signals from a diesel engine with fault and the situation that it is hard to obtain enough fualt samples,a diesel engine fault diagnosis method based on EMD and SVM was proposed.Firstly,the denoised vibration signals of three air valve clearances were decomposed into a finite number of intrinsic mode functions（IMF）,then main components were confirmed by calculating the contribution rate of variance of every IMF to original composite signal.Thirdly,an integral of main IMF components along time axis was calculated to obtain the IMF energy moment eigenvectors.Finally,with the help of SVM,the working conditions and faults of the diesel engine were classified.The results show that this method have high accuracy and good generalization even in the case of small number of samples,thus in order to get enough high correct identification rate,it is necessary to select new data as the samples when the engine speed is different.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2010年第22期2710-2714,共5页
China Mechanical Engineering
关键词
柴油机
故障诊断
能量矩
支持向量机
经验模式分解
diesel engine
fault diagnosis
energy moment
support vector machine（SVM）
empirical mode decomposition（EMD）
