摘要
电路系统在工业控制中起着极其重要的作用,随着电路越来越复杂,各电路各节点间的关系呈非线性关系,若节点发生故障,如何确定故障发生在何处成为一大难题。提出利用RBF(Radical Basis Function)神经网络可以快速地逼近任意非线性函数,且有很好的分类能力等特点,来实现对电路系统的故障分类。通过实例分析RBF可以很精确地确定电路网络中的故障来源,同时通过与BP算法比较,说明RBF在对电路故障诊断能力方面的优越性。
The circuit systems play important role in industrial control system.And these circuit behaviors are nonlinear,how to determine the fault location becomes one hard work.RBF has the capacity to quickly close any nonlinear function,and has the ability of perfect classifications.The example is test by RBF network and BP network,and the results show the advantages of RBF network.
出处
《电气传动自动化》
2009年第1期26-28,25,共4页
Electric Drive Automation
关键词
RBF神经网络
故障诊断
电路网络
BP网络
Radial Basis Function network
fault diagnosis
circuit network
BP network
