CAJ | 학술논문

针对现代储运过程管道堵塞故障诊断时，提取的过程参数多导致诊断速度慢、性能差等问题，提出了基于主成分分析（PCA）和径向基函数（RBF）神经网络故障的诊断方法。首先利用PCA方法对储运过程高维历史数据矩阵进行特征提取，提取的故障特征信息作为训练集，并给出故障特征信息的分类号；然后将其作为RBF神经网络分类器的输入输出进行故障模式识别。仿真实验表明：该方法应用于储运过程管道堵塞故障诊断，不仅大幅度地降低了诊断模型的训练时间，而且提高了诊断正确率。
침대현대저운과정관도도새고장진단시，제취적과정삼수다도치진단속도만、성능차등문제，제출료기우주성분분석（PCA）화경향기함수（RBF）신경망락고장적진단방법。수선이용PCA방법대저운과정고유역사수거구진진행특정제취，제취적고장특정신식작위훈련집，병급출고장특정신식적분류호；연후장기작위RBF신경망락분류기적수입수출진행고장모식식별。방진실험표명：해방법응용우저운과정관도도새고장진단，불부대폭도지강저료진단모형적훈련시간，이차제고료진단정학솔。
Many process parameters are selected to diagnose modern storage and transportation process blockage in pipelines, which makes diagnosis slow and its performance poor. To address it, a fault diagnosis method based on principal component analysis (PCA)and radical basis function(RBF)network was proposed. Firstly, PCA method was used to extract the feature of high histori-cal data, as the training set. And the classification of the fault characteristic information was given and then the input and output of RBF network classifiers were taken to identify the failure mode. The simulation results show that the method in pipeline of transpor-tation jam fault diagnosis not only greatly reduces the diagnostic model training time but also improves the diagnostic accuracy.