摘要
为确定合适的取样时机,以某台船用柴油机油液监测为例,采用稳健回归法制定出其原子发射光谱数据的界限值;对柴油机摩擦副磨损状况进行判断,并利用广义回归神经网络预测润滑油的剩余使用寿命;综合考虑柴油机摩擦副磨损状态和润滑油衰化情况,动态确定取样时机。
In order to determine the suitable sampling time, the oil monitoring of a marine diesel engine was studied, the limit value of atomic emission spectroscopy data was determined by robust regression. The wear state of the diesel engine was judged, and the remaining useful life of lubricating oil was predicted using generalized regression neural network. Considering the wear state of diesel engine and the decline of oil, the sampling time was determined.
出处
《润滑与密封》
CAS
CSCD
北大核心
2013年第3期102-105,共4页
Lubrication Engineering
关键词
润滑油
取样时机
稳健回归
神经网络
lubricating oil
sampling time
robust regression
neural network
