摘要
研究了将人工神经网络方法应用于热重分析数据评估润滑油某些使用性能（运动粘度和闪点）的可行性，给出了组建最优人工神经网络的规则。应用5－3－3结构BP模型人工神经网络评估润滑油的使用性能，取得了满意的结果，表明将人工神经网络应用于润滑油使用性能的评估是可行的。
The feasibility of using artificial neural network to thermogravimetric data for evaluating some physico-chemical properties (kinematic viscosity and flash point) of lubricating-oils were investigated. Some principles for establishing optimum artificial neural network are presented. A BP artificial neural network with 5 -3 -3 structure has been applied to the etaluation of the service performance of a lubricating-oil with satisfaction.
出处
《分析仪器》
CAS
1998年第4期31-34,共4页
Analytical Instrumentation
