基于RBFN故障检测的集散控制系统容错控制

DCS Fault-tolerant control Based on RBFN Fault Detection

集散控制系统(DCS)的容错控制需要解决系统的实时故障检测和容错控制律重构二大问题。本文运用径向基函数神经网络(RBFN)构造DCS的故障诊断器,通过RBFN进行容错故障范例检索,实施对DCS的控制律分段容错实时重构。这一方案在输煤控制系统中获得良好应用。

Real-time fanlt detection and control rule reconstruction are the main problems to be solved in DCS faulf-tolerant control. In this paper,the real-time segmental reconstruction of DCS control rules is realized by configurating FDD in DCS based on RBFN and retrieving fault cases of fault-tolerance. This scheme is successfully applied in coal transpotting system.