利用光谱油料分析、铁谱监测、红外光谱分析等，对８ＮＶＤ４８Ａ－２ｕ型柴油机进行了现场运行油液监测试验．根据试验结果和收集到的应用实例，总结了缸套活塞组的划伤、拉缸和腐蚀；轴承的擦伤、烧损、剥落和腐蚀；齿轮的点蚀、胶合和剥离等摩擦学故障种类及其油液监测的信息描述．利用领域专家经验给出了Ｄ－Ｓ证据组合理论的基本概率分配函数，确定了每种油液监测方法对上述摩擦学故障的支持程度，以柴油机缸套活塞组的“拉缸”故障为例，给出了Ｄ－Ｓ方法的组合支持程度的计算过程． Using oil spectrum analysis, iron spectrum monitoring, infrared spectroscopy, etc., 8NVD48A-2u diesel engine oil field monitoring test. Based on the test results and the application examples collected, the scratching, pulling cylinder and corrosion of the cylinder liner piston group were summarized; the abrasion, burning, peeling and corrosion of the bearing; the tribological failure types such as pitting, gluing and peeling of the gear And its oil monitoring information description. The basic probabilistic distribution function of D-S theory of evidence combination is given based on the field expert’s experience. The support degree of each oil monitoring method to the above tribology fault is determined. Taking “cylinder pull” of diesel cylinder liner as an example , Gives the calculation of the combined support degree of D-S method.