摘要
根据自适应谐振理论以及心理学、生物化学和神经生理学研究成果，深入研究了ＡＲＴ神经网络模型并探索出ＡＲＴ１网络、ＦＡＲＴ网络、ＡＲＴ２网络和ＡＲＴ３网络所对应的一套实用基本算法，在此基础上研制完成了ＡＲＴ神经网络集成系统软件并将其应用于轴承的状态监测和故障诊断。
The paper studies the models of ART neural networks and their basic algorithms using the reasearch achievements based on psychology,biology-chemistry and neural- physiology and adaptive resonance theory.The integrated system software of ART1,FART,ART2 and ART3 neural networks is developed.The software is applied to the real failure pattern recognition,and the result is satisfactory.
关键词
神经网络
自适应谐振
集成系统
故障诊断
s:Neural Networks
Adaptive Resonance Theory
Algorithms
Pattern Recogniton.
