摘要
基于神经网络轴系智能化故障诊断系统的开发技术 ,讨论了BP神经网络和自组织神经网络在轴系故障诊断中的应用特点 ,提出以PC为核心的扭振信号采集处理和标准样本的建立方法 ,介绍诊断系统的软件结构、模块功能、混合编程与接口技术、诊断流程和软件特点。实际应用表明 。
The shafting intelligence fault diagnosis system(SIFDS) based on NN is introduced in the paper. BP NN is compared with SOM NN. Three photo electricity sensors are installed at three different locations of the shafting to measure torsional vibration signal which are acquired and inputed to a computer. A software is developed to analyze the torsio vibration signal's FFT frequency charts and SOM NN is applied to diagnose the shafting faults. On line monitoring, off line analyzing, parameter modifying and alarm functions of this system are discussed. The technology of VB and Matlab hybrid programming and software interface are also introduced. The results of tests indicate that SIFDS based on NN is more accurate and flexible than the ordinary fault diagnosis system.
出处
《中国航海》
CSCD
北大核心
2003年第1期13-17,共5页
Navigation of China
基金
浙江省教育厅科研基金项目 (2 0 0 10 2 5 0 )
宁波市科技攻关项目 (0 0N0 40 1)
关键词
船舶工程
舰船工程
故障诊断
研究
神经网络
扭转振动
软件
Ship, Naval vessel engineering
Fault diagnosis
Study
Neural network (NN)
Torsion vibration
Software
