南京航空航天大学学报（英文版）2002，Vol.19Issue(1)：71-76,6.
基于因子模糊化BP神经网络的磨损颗粒识别
WEAR PARTICLE CLASSIFICATION BASED ON BP NEURAL NETWORK WITH FUZZY-FACTOR
李艳军 1左洪福 1吴振锋1
作者信息
- 1. 南京航空航天大学民航学院,南京,210016,中国
- 折叠
摘要
Abstract
The program of auto-identification of wear particles is given using artificial neural network (ANN) technique, based on a set of debris morphology descriptor that describes the shape characters of wear particles. The training speed of the network with the fuzzy-factor is much faster than that of the traditional methods. For example, the speed of training the network in this paper is increased five times in Exclusive OR problem (XOR problem) than other ways, and the debris classification accuracy is more than 90% by this method, and the identification speed is very fast.关键词
磨损颗粒/BP神经网络/因子模糊化/油液监测/磨粒识别Key words
wear particles/BP neural network/fuzzy-factor/lubricating oil inspection/debris identification分类
机械制造引用本文复制引用
李艳军,左洪福,吴振锋..基于因子模糊化BP神经网络的磨损颗粒识别[J].南京航空航天大学学报（英文版）,2002,19(1):71-76,6.
