船舶柴油机在用润滑油FT-IR光谱特征信息提取与应用

Extraction and Application of FT-IR Spectrum Feature Information for In-Using Lube Oil of Marine Diesel Engine

对2台船舶主柴油机润滑油进行为期1年的监测,共采集正常衰变的油样32个,另选取实际监测中发现典型的故障油样12个,应用Avatar 360型FT-IR光谱油品分析系统对上述油样进行分析。以获取的32个油样FT-IR光谱信息为研究对象,采用主成分分析法对信息进行了提取,使原始信息从5维降低至1维,建立了特征信息描述模型,将指标界限值评价法与主成分分析法相结合,通过主成分得分计算,按照正常、警告和换油3种状态,将油样进行有效的分类,实现了润滑油油品状态的评价。对获取的12个典型故障油样FT-IR光谱和原子发射光谱数据分析,表明FT-IR光谱信息能有效监测和诊断润滑油进水和燃油稀释2类故障现象,同时从微观上揭示2类现象对润滑油性能的影响,并能早于原子发射光谱发现设备摩擦学系统故障源。

In order to study how to apply the Fourier Transform Infrared （FT-IR） spectrum informa tion to monitor marine diesel engine tribology system,two engines on board were monitored for one year. 32 normal used oil samples taken from them and other 12 typical fault oil samples taken in daily work were analyzed by FT-IR spectrometer. Using Principal Component Analysis （PCA） Method to extract FT-IR spectrum information samples, the dimension number of the original information was reduced from five to one to set up the feature information indication model. The 32 samples were evaluated according to the scores achieved based on Index Limit Value Assessing Method and PCA Method and classified as 3 kinds of states, normal, alarm and to be replaced. By analyzing the FT-IR spectrum and emission spectroscopy data of 12 typical fault samples, it is exhibited that FT-IR spectrum information can efficiently monitor and diagnose faults of water contamination and fuel contamination, and reveal in microcosmic level the influence of above two kinds of faults on quality of the in-using lubricant, so as to find the fxult source in engine tribological system earlier than from emission spectroscopy data and remove it in time.