摘要
本文提出了一种基于主成分分析(PCA)和概率神经网络(PNN)的故障诊断方法。该方法首先利用PCA分析建模消除变量之间的非线性关联,降低噪声的影响,在保证数据信息丢失最少的情况下,大大降低了原始数据空间的维数,然后利用概率神经网络对降维后的数据进行模式分类,最后结合某汽车发动机的故障诊断进行仿真研究。仿真结果表明,该方法是有效可行的。
In this paper,a fault diagnosis method based on principal component analysis（PCA）and probabilistic neural network（PNN）is proposed.Firstly,As an analyzing and modelling tool,PCA is introduced to eliminate nonlinear combination of the variables and decrease the influence of the noise,and it also decreases dimensions of the original variables under the condition that the missing of the data information is least.Then,the PNN is used to recognize the reduced data.Finally,this method is combined with the study of automobile engine fault diagnosis.The simulation result indicates that the method mentioned above is effective and feasible.
出处
《网络安全技术与应用》
2010年第6期58-60,共3页
Network Security Technology & Application
关键词
主成分分析
概率神经网络
故障检测
故障诊断
Principal Component Analysis
Probabilistic Neural Network
Fault Detection
Fault Diagnosis
