摘要
介绍一种可自动从实例中获取知识的、具有高模式聚类精度的新型神经网络———EBL网(examplebasedlearningnetwork),及其在故障诊断中的应用.研究结果表明:EBL模式聚类系统可逐个从实例中学习知识,并可同时给出模式识别结果及集群的信息.EBL模式聚类系统适用于实时故障诊断.
This paper discusses a new type of neural network called EBL network (example based learning network) and how to apply it to fault diagnosis.The research results have shown that the EBL pattern clustering system can learn and accumulate knowledge from each example.Moreover both the results of the pattern recognition and the information of the clusters stored in the system can be given.The EBL pattern clustering system is suitable for fault diagnosis with real time demand.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1999年第8期5-8,4,共5页
Acta Electronica Sinica
基金
国家教委留学回国人员科研启动基金
关键词
神经网络
实例学习
故障诊断
化工过程
Neural network,Fault diagnosis,Example learning,Pattern clustering