基于PCA神经网络和D-S决策的瓦斯传感器故障辨识  

Fault Identification for Gas Sensor Based on PCA-neural Network and D-S Evidence Theory

作　　者：黄丹[1] 徐平安[1] 王其军[1,2] 任玉东[1] 严彬[1] 

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001 [2]淮南职业技术学院,安徽淮南232001

出　　处：《仪表技术与传感器》2015年第3期99-103,共5页Instrument Technique and Sensor

基　　金：国家自然科学基金青年科学基金项目(51304007)

摘　　要：针对瓦斯传感器故障诊断时,存在提取的样本数据空间维数大、诊断实时性差、诊断结论的识别能力低和存在不确定性的问题,提出了一种基于主元分析(PCA)-神经网络和D-S证据理论集成的故障诊断策略。使用主元分析方法对高维故障样本空间数据进行降维,再结合神经网络分类器进行故障模式识别。并且运用DS证据理论对神经网络分类器的故障诊断结果进行数据融合。仿真实验表明:该诊断方法改善了神经网络对瓦斯传感器故障诊断准确率的同时提高了诊断速度,并且降低了故障结论的不确定性以及提高了结论的识别与决策能力。For the problems existing in the gas sensor fault diagnosis such as the large space dimension of the sample data,weak real-time of fault diagnosis,poor identification ability of the diagnosis result and the uncertainty,fault diagnosis strategy was proposed based on principal component analysis（ PCA） neural network and D-S evidence theory. The principal component analysis（ PCA） was used to reduce the high dimension of the fault sample space data,combining the neural network classifiers to identify the fault mode,and the DS evidence theory was used for data fusion in the fault diagnosis results of the neural network classifiers.The simulation results show that the accuracy rate can be improved and the diagnosis speed can be increased by the use of the method. Furthermore,the uncertainty of fault conclusion can be reduced and the ability of the conclusion recognition and decisionmaking can be improved.

关 键 词：瓦斯传感器 故障诊断 主成分分析 神经网络 DS证据理论 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]