摘要
化工生产过程一般都非常复杂,控制回路与测控参数很多,生产过程的故障检测与诊断问题非常困难。文中提出一种基于神经网络的多级故障诊断系统。采用三级递阶模糊神经网络,降解整个系统故障诊断问题的复杂性,同时采用所有子神经网络全局并行的推理方式,具有快速处理能力,适合系统实时在线故障诊断。
In jeneral,chemical processes in industry are extremely complicated.It is very difficult to detect and diagnose the faults in industrial processes.This paper presents a fault diagnostic system based on multilevel neural networks.By using a 3level neural network,the complication of the fault diagnosis in a real system is simplified and it also has the quick handling ability to match the online fault diagnosis by using the rational measurement of juxtaposing subneural networks.
出处
《工业仪表与自动化装置》
2003年第1期37-39,40,共4页
Industrial Instrumentation & Automation
