基于功率谱包络能量和SVM的舰用发动机故障诊断

Based on Power Spectral Envelope Energy and SVM Warship Engine Fault Diagnosis

摘要:

发动机是军舰上的重要部件之一,其稳定性对军舰的正常航行具有重要影响。以舰用发动机关键部件(主泵轴承)为具体研究对象,提出了基于功率谱包络能量和支持向量机相结合的故障诊断方法。首先获取了大量可表征舰用发动机主泵轴承健康状态的振动加速度信息,对其进行功率谱分析,获得其功率谱的包络能量；以获取的舰用发动机主泵轴承功率谱的包络能量构建特征向量,并设计基于SVM的舰用发动机主泵轴承故障诊断模型,对主泵轴承的故障进行诊断研究。研究结果表明,采用基于功率谱包络能量和SVM相结合的舰用发动机关键部件故障诊断方法,可以很好实现主泵轴承的故障诊断效能,为舰用发动机主泵轴承故障诊断的工程应用奠定了基础。

Abstract:

The engine is one of the important components warship, its stability has a significant impact on the normal navigation of warships. The key components of ship engines (main pump bearings) the research object proposed network fault diagnosis method based on support vector machines for energy and power spectral envelope. First, get a lot can be characterized warship bearing the health status of the main pump engine vibration acceleration information, its power spectrum analysis, power spectrum obtained envelope of energy. The acquired warship bearing the main pump engine power spectral envelope energy feature vectors constructed and designed SVM warships engine main pump bearing fault diagnosis model based on the primary pump bearing fault diagnosis. The results show that, based on critical engine components warships network fault diagnosis SVM combination of energy and power spectrum package, you can achieve a good diagnostic performance of the main pump bearing fault laid engines for ship main pump engineering bearing fault diagnosis basis.