基于粒计算-神经网络的故障诊断方法
首发时间:2009-03-03
摘要:本文结合粒计算和神经网络的优点,提出一种基于粒计算-神经网络(GrC-NN)的智能故障诊断方法。对采集来的故障样本进行预处理和离散化形成决策表,利用二进制粒矩阵的运算来约简决策表以得到最小属性集;根据约简后的最优决策规则,建立基于RBF神经网络的故障诊断系统。通过水电机组的故障诊断实例,表明了GrC-NN诊断方法的有效性。
For information in English, please click here
Fault Diagnosis Method Based on Granular Computing and Neural Network
Abstract:A fault diagnosis method using granular computing and artificial neural network is developed, considering that neural networks are best for solving nonlinear problems while granular computing is good for data reduction. Binary granular matrix is applied to the decision table reduction to find minimum attribute set under the same classification ability. The reduced system is utilized to the neural fault classifier, which indicates that granular-computing-based-reduction reduces the dimension of input to neural network and improves the efficiency of training. An example applying to the hydro generator unit shows the effectiveness of the GrC-NN method proposed in the paper.
Keywords: granular computing binary granular matrix neural network fault diagnosis
论文图表:
引用
No.2984243039912360****
同行评议
共计0人参与
勘误表
基于粒计算-神经网络的故障诊断方法
评论
全部评论0/1000